CM SP PHYv3.1 I01 131029

Data-Over-Cable Service Interface Specifications
DOCSIS 3.1
Physical Layer Specification

CM-SP-PHYv3.1-I01-131029
ISSUED
Notice
This DOCSIS specification is the result of a cooperative effort
undertaken at the direction of Cable Television Laboratories, Inc. for the
benefit of the cable industry and its customers. This document may
contain references to other documents not owned or controlled by
CableLabs . Use and understanding of this document may require

access to such other documents. Designing, manufacturing, distributing,
using, selling, or servicing products, or providing services, based on this
document may require intellectual property licenses from third parties
for technology referenced in this document.
Neither CableLabs nor any member company is responsible to any
party for any liability of any nature whatsoever resulting from or arising
out of use or reliance upon this document, or any document referenced
herein. This document is furnished on an "AS IS" basis and neither
CableLabs nor its members provides any representation or warranty,
express or implied, regarding the accuracy, completeness,
noninfringement, or fitness for a particular purpose of this document, or
any document referenced herein.
Cable Television Laboratories, Inc. 2013
CM-SP-PHYv3.1-I01-131029
DISCLAIMER
This document is published by Cable Television Laboratories, Inc. ("CableLabs").
CableLabs reserves the right to revise this document for any reason including, but not limited to, changes in laws,
regulations, or standards promulgated by various agencies; technological advances; or changes in equipment design,
manufacturing techniques, or operating procedures described, or referred to, herein. CableLabs makes no
representation or warranty, express or implied, with respect to the completeness, accuracy, or utility of the document
or any information or opinion contained in the report. Any use or reliance on the information or opinion is at the risk
of the user, and CableLabs shall not be liable for any damage or injury incurred by any person arising out of the
completeness, accuracy, or utility of any information or opinion contained in the document.
This document is not to be construed to suggest that any affiliated company modify or change any of its products or
procedures, nor does this document represent a commitment by CableLabs or any cable member to purchase any
product whether or not it meets the described characteristics. Nothing contained herein shall be construed to confer
any license or right to any intellectual property, whether or not the use of any information herein necessarily utilizes
such intellectual property. This document is not to be construed as an endorsement of any product or company or as
the adoption or promulgation of any guidelines, standards, or recommendations.
ii
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Document Status Sheet

Document Control Number:
Document Title:
Revision History:
Date:
Status:
Distribution Restrictions:
CM-SP-PHYv3.1-I01-131029
I01 - Released 10/29/13
October 29, 2013
Work in
Progress
Draft
Issued
Closed
Author Only
CL/Member
CL/ Member/
Vendor
Public
Key to Document Status Codes

Work in Progress
An incomplete document, designed to guide discussion and generate feedback

that may include several alternative requirements for consideration.
Draft
A document in specification format considered largely complete, but lacking

review by Members and vendors. Drafts are susceptible to substantial change
during the review process.
Issued
A generally public document that has undergone Member and Technology

Supplier review, cross-vendor interoperability, and is for Certification testing if
applicable. Issued Specifications are subject to the Engineering Change Process.
Closed
A static document, reviewed, tested, validated, and closed to further engineering

changes.
Trademarks
CableLabs is a registered trademark of Cable Television Laboratories, Inc. Other CableLabs marks are listed at
http://www.cablelabs.com/certqual/trademarks. All other marks are the property of their respective owners.
10/29/13
CableLabs
iii
CM-SP-PHYv3.1-I01-131029
Contents
1
SCOPE ............................................................................................................................................................ 1
1.1
INTRODUCTION AND PURPOSE...............................................................................................................................1
1.2
BACKGROUND ....................................................................................................................................................1
1.2.1
Broadband Access Network ...................................................................................................................1
1.2.2
Network and System Architecture .........................................................................................................2
1.2.3
Service Goals ..........................................................................................................................................3
1.2.4
Statement of Compatibility ....................................................................................................................3
1.2.5
Reference Architecture ..........................................................................................................................4
1.2.6
DOCSIS 3.1 Documents...........................................................................................................................4
1.3
REQUIREMENTS ..................................................................................................................................................5
1.4
CONVENTIONS ...................................................................................................................................................5
1.5
ORGANIZATION OF DOCUMENT .............................................................................................................................6
REFERENCES ................................................................................................................................................... 7
2.1
2.2
2.3
NORMATIVE REFERENCES .....................................................................................................................................7

INFORMATIVE REFERENCES ...................................................................................................................................8
REFERENCE ACQUISITION .....................................................................................................................................8
TERMS AND DEFINITIONS............................................................................................................................... 9
ABBREVIATIONS AND ACRONYMS ............................................................................................................... 21
OVERVIEW AND FUNCTIONAL ASSUMPTIONS .............................................................................................. 26

5.1
OVERVIEW ......................................................................................................................................................26
5.2
FUNCTIONAL ASSUMPTIONS................................................................................................................................27
5.2.1
Equipment Assumptions ......................................................................................................................27
5.2.2
RF Channel Assumptions ......................................................................................................................28
5.2.3
Transmission Levels..............................................................................................................................30
5.2.4
Frequency Inversion .............................................................................................................................30
PHY SUBLAYER FOR SC-QAM ........................................................................................................................ 31

6.1
SCOPE ............................................................................................................................................................31
6.2
UPSTREAM TRANSMIT AND RECEIVE .....................................................................................................................31
6.2.1
Overview ..............................................................................................................................................31
6.2.2
Signal Processing Requirements ..........................................................................................................31
6.2.3
Modulation Formats ............................................................................................................................31
6.2.4
R-S Encode ...........................................................................................................................................31
6.2.5
Upstream R-S Frame Structure (Multiple Transmit Enabled)...............................................................31
6.2.6
Upstream R-S Frame Structure (Multiple Transmit Disabled) ..............................................................31
6.2.7
TDMA Byte Interleaver .........................................................................................................................32
6.2.8
Scrambler (randomizer) .......................................................................................................................32
6.2.9
TCM Encoder ........................................................................................................................................32
6.2.10 Preamble Prepend................................................................................................................................32
6.2.11 Modulation Rates.................................................................................................................................32
6.2.12 S-CDMA Framer and Interleaver ..........................................................................................................32
6.2.13 S-CDMA Framer ...................................................................................................................................32
6.2.14 Symbol Mapping ..................................................................................................................................32
6.2.15 S-CDMA Spreader .................................................................................................................................32
6.2.16 Transmit Pre-Equalizer .........................................................................................................................32
iv
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
6.2.17 Spectral Shaping ..................................................................................................................................32

6.2.18 Relative Processing Delays ...................................................................................................................32
6.2.19 Transmit Power Requirements .............................................................................................................32
6.2.20 Burst Profiles ........................................................................................................................................32
6.2.21 Burst Timing Convention ......................................................................................................................32
6.2.22 Fidelity Requirements ..........................................................................................................................32
6.2.23 Upstream Demodulator Input Power Characteristics ..........................................................................33
6.2.24 Upstream Electrical Output from the CM ............................................................................................33
6.2.25 Upstream CM Transmitter Capabilities................................................................................................33
6.3
DOWNSTREAM TRANSMIT ..................................................................................................................................33
6.3.1
Downstream Protocol ..........................................................................................................................33
6.3.2
Spectrum Format .................................................................................................................................33
6.3.3
Scaleable Interleaving to Support Video and High-Speed Data Services .............................................33
6.3.4
Downstream Frequency Plan ...............................................................................................................33
6.3.5
DRFI Output Electrical ..........................................................................................................................33
6.3.6
CMTS or EQAM Clock Generation ........................................................................................................33
6.3.7
Downstream Symbol Clock Jitter for Synchronous Operation ..............................................................34
6.3.8
Downstream Symbol Clock Drift for Synchronous Operation ..............................................................34
6.3.9
Timestamp Jitter ..................................................................................................................................34
6.4
DOWNSTREAM RECEIVE .....................................................................................................................................34
6.4.1
Downstream Protocol and Interleaving Support..................................................................................34
6.4.2
Downstream Electrical Input to the CM ...............................................................................................34
6.4.3
CM BER Performance ...........................................................................................................................34
6.4.4
Downstream Multiple Receiver Capabilities ........................................................................................34
6.4.5
Non-Synchronous DS Channel Support ................................................................................................34
7
PHY SUBLAYER FOR OFDM ........................................................................................................................... 35

7.1
SCOPE ............................................................................................................................................................35
7.2
UPSTREAM AND DOWNSTREAM FREQUENCY PLAN ..................................................................................................35
7.2.1
Downstream CM Spectrum ..................................................................................................................35
7.2.2
Downstream CMTS Spectrum ..............................................................................................................35
7.2.3
Upstream CM Spectrum .......................................................................................................................36
7.2.4
Upstream CMTS Spectrum ...................................................................................................................36
7.2.5
Exclusion Band Rules ............................................................................................................................36
7.3
OFDM NUMEROLOGY ......................................................................................................................................37
7.3.1
Downstream OFDM Numerology.........................................................................................................37
7.3.2
Upstream OFDMA Numerology ...........................................................................................................38
7.3.3
Subcarrier Clocking ..............................................................................................................................39
7.4
UPSTREAM TRANSMIT AND RECEIVE .....................................................................................................................39
7.4.1
Overview ..............................................................................................................................................39
7.4.2
Signal Processing and Requirements ...................................................................................................41
7.4.3
Time and Frequency Synchronization ..................................................................................................43
7.4.4
Forward Error Correction .....................................................................................................................44
7.4.5
Data Randomization ............................................................................................................................50
7.4.6
Time and Frequency Interleaving and De-interleaving ........................................................................50
7.4.7
Mapping of Bits to Cell Words .............................................................................................................57
7.4.8
Mapping and De-mapping Bits to/from QAM subcarriers ...................................................................58
7.4.9
REQ Messages ......................................................................................................................................58
7.4.10 IDFT ......................................................................................................................................................59
7.4.11 Cyclic Prefix and Windowing ................................................................................................................60
7.4.12 Fidelity Requirements ..........................................................................................................................66
7.4.13 Cable Modem Transmitter Output Requirements................................................................................74
7.4.14 CMTS Receiver Input Power Requirements ..........................................................................................75
10/29/13
CableLabs
CM-SP-PHYv3.1-I01-131029
7.4.15 Ranging ................................................................................................................................................76

7.4.16 Upstream Pilot Structure .....................................................................................................................87
7.4.17 Upstream Pre-Equalization ..................................................................................................................93
7.5
DOWNSTREAM TRANSMIT AND RECEIVE................................................................................................................93
7.5.1
Overview ..............................................................................................................................................93
7.5.2
Signal Processing .................................................................................................................................95
7.5.3
Time and Frequency Synchronization ..................................................................................................98
7.5.4
Downstream Forward Error Correction................................................................................................ 99
7.5.5
Mapping Bits to QAM Constellations .................................................................................................107
7.5.6
Interleaving and De-interleaving .......................................................................................................117
7.5.7
IDFT ....................................................................................................................................................122
7.5.8
Cyclic Prefix and Windowing ..............................................................................................................124
7.5.9
Fidelity Requirements ........................................................................................................................125
7.5.10 CMTS Transmitter Output Requirements ...........................................................................................133
7.5.11 Cable Modem Receiver Input Requirements ......................................................................................136
7.5.12 Cable Modem Receiver Capabilities ...................................................................................................138
7.5.13 Physical Layer Link Channel (PLC) ......................................................................................................139
7.5.14 Next Codeword Pointer ......................................................................................................................150
7.5.15 Downstream Pilot Patterns ................................................................................................................155
8
PHYMAC CONVERGENCE .......................................................................................................................... 166

8.1
SCOPE ..........................................................................................................................................................166
8.2
UPSTREAM ....................................................................................................................................................167
8.2.1
Profiles ...............................................................................................................................................167
8.2.2
Minislots.............................................................................................................................................167
8.2.3
Subslots ..............................................................................................................................................169
8.3
DOWNSTREAM ...............................................................................................................................................171
8.3.1
Operation ...........................................................................................................................................171
8.3.2
MAC Frame to Codewords .................................................................................................................173
8.3.3
Subcarrier Numbering Conventions ...................................................................................................174
8.3.4
Next Codeword Pointer ......................................................................................................................174
PROACTIVE NETWORK MAINTENANCE....................................................................................................... 179

9.1
SCOPE ..........................................................................................................................................................179
9.2
SYSTEM DESCRIPTION ......................................................................................................................................179
9.3
DOWNSTREAM PNM REQUIREMENTS ................................................................................................................180
9.3.1
Downstream Symbol Capture ............................................................................................................180
9.3.2
Downstream Wideband Spectrum Capture .......................................................................................180
9.3.3
Downstream Noise Power Ratio (NPR) Measurement .......................................................................181
9.3.4
Downstream Equalizer Coefficients ...................................................................................................181
9.3.5
Downstream Constellation Display ....................................................................................................181
9.3.6
Downstream Receive Modulation Error Ratio (RxMER) Per Subcarrier .............................................181
9.3.7
Downstream FEC Statistics ................................................................................................................182
9.3.8
Downstream Histogram .....................................................................................................................183
9.3.9
Downstream Profile Performance Metrics .........................................................................................183
9.4
UPSTREAM PNM REQUIREMENTS .....................................................................................................................183
9.4.1
Upstream Capture for Quiet Period and Probe Symbol .....................................................................183
9.4.2
Upstream Triggered Spectrum Capture .............................................................................................184
9.4.3
Upstream Impulse Noise Statistics .....................................................................................................184
9.4.4
Upstream Equalizer Coefficients ........................................................................................................184
9.4.5
Upstream FEC Statistics .....................................................................................................................184
9.4.6
Upstream Histogram .........................................................................................................................185
vi
CableLabs
10/29/13

ANNEX A
A.1
A.2
CM-SP-PHYv3.1-I01-131029
QAM CONSTELLATION MAPPINGS .................................................................................................. 186

QAM CONSTELLATIONS ...................................................................................................................................186
QAM CONSTELLATION SCALING ........................................................................................................................196
ANNEX B
RFOG OPERATING MODE ................................................................................................................ 197
ANNEX C
ADDITIONS AND MODIFICATIONS FOR EUROPEAN SPECIFICATION WITH SC-QAM OPERATION ..... 198
ANNEX D
ADDITIONS AND MODIFICATIONS FOR CHINESE SPECIFICATION WITH SC-QAM OPERATION ......... 199
ANNEX E
24-BIT CYCLIC REDUNDANCY CHECK (CRC) CODE ............................................................................ 200
Figures
FIGURE 11 - THE DOCSIS NETWORK ...................................................................................................................................2
FIGURE 12 - TRANSPARENT IP TRAFFIC THROUGH THE DATA-OVER-CABLE SYSTEM ......................................................................3
FIGURE 1-3 - DATA-OVER-CABLE REFERENCE ARCHITECTURE .....................................................................................................4
FIGURE 71 - OFDMA FRAME STRUCTURE ..........................................................................................................................41
FIGURE 72 - UPSTREAM TRANSMITTER BLOCK DIAGRAM ........................................................................................................41
FIGURE 73 - UPSTREAM TRANSMITTER BLOCK DIAGRAM .......................................................................................................42
FIGURE 74 - UPSTREAM DATA RANDOMIZER .......................................................................................................................50
FIGURE 75 - CALCULATING NUMBER OF MINISLOTS IN EACH BLOCK FOR UPSTREAM INTERLEAVING ...............................................51
FIGURE 76 - ILLUSTRATING MINISLOTS OF A GRANT OVER WHICH INTERLEAVING IS PERFORMED....................................................52
FIGURE 77 - SAMPLE INTERLEAVER BLOCK DIAGRAM ............................................................................................................53
FIGURE 78 - INTERLEAVING A GRANT WITHIN AN OFDMA FRAME ..........................................................................................55
FIGURE 79 - BIT-REVERSED COUNTER IMPLEMENTATION .......................................................................................................57
FIGURE 710 - BITSTREAM TO QAM M-TUPLE MAPPING .......................................................................................................58
FIGURE 711 - REQ MESSAGES PROCESSING........................................................................................................................59
FIGURE 712 - INVERSE DISCRETE FOURIER TRANSFORM.........................................................................................................60
FIGURE 713 - SIGNAL WITH MICRO-REFLECTION AT THE RECEIVER ..........................................................................................60
FIGURE 714 - CYCLIC PREFIX AND WINDOWING ALGORITHM ..................................................................................................63
FIGURE 715 - TAPERING WINDOW ....................................................................................................................................64
FIGURE 716 - RANGING STEPS ..........................................................................................................................................76
FIGURE 717 - INITIAL RANGING ZONE ................................................................................................................................77
FIGURE 718 - INITIAL RANGING SIGNAL ..............................................................................................................................77
FIGURE 719 - INITIAL RANGING ADMISSION SLOT STRUCTURE ................................................................................................78
FIGURE 720 - BLOCK DIAGRAM OF INITIAL RANGING TRANSMITTER PROCESSING .......................................................................78
FIGURE 721 - LDPC TWO-PERIOD PUNCTURING ENCODER FOR INITIAL RANGING FEC ...............................................................79
FIGURE 722 - INITIAL RANGING SYMBOL PAIR STRUCTURE .....................................................................................................80
FIGURE 723 - FINE RANGING SIGNAL .................................................................................................................................81
FIGURE 724 - FINE RANGING SIGNAL TRANSMISSION ............................................................................................................82
FIGURE 725 - FINE RANGING TRANSMITTER PROCESSING ......................................................................................................82
FIGURE 726 - SHORTENING AND PUNCTURING ENCODER FOR THE FINE RANGING FEC................................................................83
FIGURE 727 - FINE RANGING AND EXCLUSION BANDS ...........................................................................................................84
FIGURE 728 - POLYNOMIAL SEQUENCE FOR PSEUDORANDOM BINARY SEQUENCE GENERATION ....................................................85
FIGURE 729 - 4K FFT EXAMPLE, ALL SUBCARRIERS USED FOR PROBING, NO SKIPPING ................................................................86
FIGURE 730 - 4K FFT EXAMPLE, ALTERNATE SUBCARRIERS USED FOR PROBING ........................................................................86
FIGURE 731 - EDGE AND BODY MINISLOTS IN A TRANSMISSION BURST.....................................................................................87
FIGURE 732 - PILOT PATTERNS 1-4 FOR MINISLOTS WITH 8 SUBCARRIERS ................................................................................88
FIGURE 733 - PILOT PATTERNS 5-7 FOR MINISLOTS WITH 8 SUBCARRIERS ................................................................................89
FIGURE 734 - PILOT PATTERNS 8-11 FOR MINISLOTS WITH 16 SUBCARRIERS ............................................................................90
FIGURE 735 - PILOT PATTERNS 12-14 FOR MINISLOTS WITH 16 SUBCARRIERS ..........................................................................91
FIGURE 736 - PILOT PATTERN FOR SUBSLOTS WITH 8 SUBCARRIERS .........................................................................................92
10/29/13
CableLabs
vii
CM-SP-PHYv3.1-I01-131029
FIGURE 737 - PILOT PATTERN FOR SUBSLOTS WITH 16 SUBCARRIERS .......................................................................................92

FIGURE 738 - DOWNSTREAM PHY PROCESSING...................................................................................................................96
FIGURE 739 - CODEWORD SHORTENING PROCESS ..............................................................................................................102
FIGURE 740 - PADDING PROCESS ....................................................................................................................................103
FIGURE 741 - BIT DE-INTERLEAVER BLOCK FOR A SHORTENED CODEWORD..............................................................................106
FIGURE 742 - FEC DECODING PROCESS............................................................................................................................107
FIGURE 743 - BITS TO QAM CONSTELLATION MAPPING......................................................................................................108
FIGURE 744 - BIT-TO-CELL WORD DEMULTIPLEXER ............................................................................................................109
FIGURE 745 - LINEAR FEEDBACK SHIFT REGISTER FOR RANDOMIZATION SEQUENCE ..................................................................112
FIGURE 746 - BIT LOADING, SYMBOL MAPPING, AND INTERLEAVING .....................................................................................114
FIGURE 747 - NCP INSERTION ........................................................................................................................................117
FIGURE 748 - TIME INTERLEAVER STRUCTURE ....................................................................................................................118
FIGURE 749 - TWO-DIMENSIONAL BLOCK STRUCTURE ........................................................................................................119
FIGURE 750 - LINEAR FEEDBACK SHIFT REGISTER................................................................................................................120
FIGURE 751 - FREQUENCY INTERLEAVER ROTATION DEFINITION ............................................................................................120
FIGURE 752 - INVERSE DISCRETE FOURIER TRANSFORM .......................................................................................................123
FIGURE 753 - STRUCTURE OF THE PLC .............................................................................................................................139
FIGURE 754 - EXAMPLES OF PLC PLACEMENT ....................................................................................................................141
FIGURE 755 - PHYSICAL LAYER OPERATIONS FOR FORMING THE PLC SUBCARRIERS...................................................................142
FIGURE 756 - MAPPING BYTES INTO A BIT STREAM FOR FEC ENCODING ................................................................................144
FIGURE 757 - PUNCTURING ENCODER FOR THE PLC FEC .....................................................................................................145
FIGURE 758 - MAPPING ENCODED BIT STREAM INTO A STREAM OF NIBBLES ...........................................................................145
FIGURE 759 - BLOCK INTERLEAVING OF PLC SUBCARRIERS FOR 4K FFT ..................................................................................146
FIGURE 760 - BLOCK INTERLEAVING OF PLC SUBCARRIERS FOR 8K FFT ..................................................................................147
FIGURE 761 - LINEAR FEEDBACK SHIFT REGISTER FOR PLC RANDOMIZATION ...........................................................................148
FIGURE 762 - TIME - FREQUENCY PLANE REPRESENTATION OF PLC TIMESTAMP SYNCHRONIZATION ............................................149
FIGURE 763 - TIME DOMAIN REPRESENTATION OF PLC TIMESTAMP SYNCHRONIZATION............................................................150
FIGURE 764 - MAPPING NCP BYTES INTO A BIT STREAM FOR FEC ENCODING .........................................................................151
FIGURE 765 - SHORTENING AND PUNCTURING ENCODER FOR THE NCP FEC ...........................................................................152
FIGURE 766 - 64-QAM CONSTELLATION MAPPING OF {, , , , , , } .....................................................153
FIGURE 767 - 16QAM CONSTELLATION MAPPING OF {, , , , } ....................................................................154
FIGURE 768 - QPSK CONSTELLATION MAPPING OF {, , } .........................................................................................154
FIGURE 769 - 4K FFT DOWNSTREAM PILOT PATTERN .........................................................................................................156
FIGURE 770 - A DOWNSTREAM SCATTERED PILOT PATTERN FOR 8K FFT (FOR EXPLANATION PURPOSES ONLY) .............................158
FIGURE 771 - 8K FFT DOWNSTREAM SCATTERED PILOT PATTERN .........................................................................................160
FIGURE 772 - PLACEMENT OF PREDEFINED CONTINUOUS PILOTS AROUND THE PLC .................................................................162
FIGURE 773 - 13-BIT LINEAR FEEDBACK SHIFT REGISTER FOR THE PILOT MODULATION PSEUDO-RANDOM SEQUENCE ....................164
FIGURE 81 - MINISLOT DATA BIT ORDERING .....................................................................................................................169
FIGURE 82 - SUBSLOT STRUCTURE ...................................................................................................................................170
FIGURE 83 - DATA MAPPING FOR A 4X8 SUBSLOT ..............................................................................................................170
FIGURE 84 - DATA MAPPING FOR A 2X16 SUBSLOT ............................................................................................................171
FIGURE 85 - DOWNSTREAM CONVERGENCE LAYER BLOCK DIAGRAM ......................................................................................172
FIGURE 86 - DOCSIS FRAME TO CODEWORD MAPPING ......................................................................................................173
FIGURE 87 - DATA AND NCP PRIOR TO INTERLEAVING.........................................................................................................174
FIGURE 88 - NCP MESSAGE BLOCK .................................................................................................................................175
FIGURE 89 - NCP EXAMPLES ..........................................................................................................................................177
FIGURE 91 - TEST POINTS IN CM AND CMTS SUPPORTING PROACTIVE NETWORK MAINTENANCE ...............................................180
FIGURE A1 - BPSK CONSTELLATION MAPPING OF {}......................................................................................................186
FIGURE A2 - QPSK CONSTELLATION MAPPING OF {} ................................................................................................186
FIGURE A3 - 8-QAM CONSTELLATION MAPPING OF { } .......................................................................................187
FIGURE A4 - 16-QAM CONSTELLATION MAPPING OF { } ................................................................................187
FIGURE A5 - 32-QAM CONSTELLATION MAPPING OF {} .............................................................................187
FIGURE A6 - 64-QAM CONSTELLATION MAPPING OF {} ........................................................................188
viii
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
FIGURE A7 - 128-QAM CONSTELLATION MAPPING OF {} ..................................................................188

FIGURE A8 - MAPPING OF BITS { } OF { , , } FOR CONSTELLATIONS WITH ONLY ONE QUADRANT
DEFINED ...............................................................................................................................................................189
FIGURE A9 - REFLECTIVE MAPPING OF BITS { , , } FOR ALL CONSTELLATIONS (EXCEPT BPSK) ............................189
FIGURE A10 - 256-QAM CONSTELLATION MAPPING OF {} ON TO QUADRANT 1 ........................................190
FIGURE A11 - 512-QAM CONSTELLATION MAPPING OF {} ON TO QUADRANT 1 ...................................190
FIGURE A12 - 1024-QAM CONSTELLATION MAPPING OF {} ON TO QUADRANT 1.............................191
FIGURE A13 - 2048-QAM CONSTELLATION MAPPING OF {} ON TO QUADRANT 1 .....................192
FIGURE A14 - 4096-QAM CONSTELLATION MAPPING OF {} ON TO QUADRANT 1 ..............193
FIGURE A15 - 8192-QAM CONSTELLATION MAPPING OF {} ON TO
QUADRANT 1.........................................................................................................................................................194
FIGURE A16 - 16384-QAM CONSTELLATION MAPPING OF {} ON TO
QUADRANT 1.........................................................................................................................................................195
Tables
TABLE 11 - DOCSIS 3.1 SERIES OF SPECIFICATIONS ................................................................................................................4
TABLE 12 - DOCSIS 3.1 RELATED SPECIFICATIONS .................................................................................................................5
TABLE 51 - TYPICAL DOWNSTREAM RF CHANNEL TRANSMISSION CHARACTERISTICS....................................................................28
TABLE 52 - TYPICAL UPSTREAM RF CHANNEL TRANSMISSION CHARACTERISTICS.........................................................................29
TABLE 71 - DOWNSTREAM OFDM PARAMETERS .................................................................................................................38
TABLE 72 - UPSTREAM OFDMA PARAMETERS ....................................................................................................................38
TABLE 73 - UPSTREAM CODEWORD PARAMETERS ................................................................................................................44
TABLE 74 - CYCLIC PREFIX (CP) VALUES .............................................................................................................................61
TABLE 75 - ROLL-OFF PREFIX (RP) VALUES .........................................................................................................................61
TABLE 76 - SPURIOUS EMISSIONS ......................................................................................................................................68
TABLE 77 - SPURIOUS EMISSIONS REQUIREMENTS IN THE UPSTREAM FREQUENCY RANGE FOR GRANTS OF UNDER-GRANT
1
HOLD BANDWIDTH AND LARGER ................................................................................................................................70
TABLE 78 - ADJACENT CHANNEL SPURIOUS EMISSIONS REQUIREMENTS RELATIVE TO THE PER CHANNEL TRANSMITTED BURST POWER
LEVEL FOR EACH CHANNEL .........................................................................................................................................72
TABLE 79 - UPSTREAM MER REQUIREMENTS (WITH PRE-EQUALIZATION) ................................................................................74
TABLE 710 - UPSTREAM MER REQUIREMENTS (NO PRE-EQUALIZATION) .................................................................................74
TABLE 711 - CM TRANSMITTER OUTPUT SIGNAL CHARACTERISTICS .........................................................................................74
TABLE 712 - UPSTREAM CHANNEL DEMODULATOR INPUT POWER CHARACTERISTICS ..................................................................75
TABLE 713 - (160,80) LDPC CODE PARITY CHECK MATRIX ...................................................................................................79
TABLE 714 - CYCLIC PREFIX AND ROLLOFF SAMPLES FOR INITIAL RANGING..............................................................................80
TABLE 715 - (480, 288) LDPC CODE PARITY CHECK MATRIX ................................................................................................83
TABLE 716 - MIXED MODULATION WITH 1.5 DB SNR GRANULARITY ....................................................................................100
TABLE 717 - CODING PARAMETERS (FOR SHORT CODEWORDS NLDPC = 16,200 AND CODE RATE 8/9) ..........................................101
TABLE 718 - BIT INTERLEAVER STRUCTURE ........................................................................................................................104
TABLE 719 - COLUMN TWISTING PARAMETER TC (COLUMNS 0 - 11) ......................................................................................104
TABLE 720 - COLUMN TWISTING PARAMETER TC (COLUMNS 12 - 23) ....................................................................................104
TABLE 725 - PARAMETERS FOR BIT-MAPPING ONTO CONSTELLATIONS ...................................................................................108
TABLE 726 - NUMBER OF SUB-STREAMS IN DEMULTIPLEXER ................................................................................................109
TABLE 727 - PARAMETERS FOR DEMULTIPLEXING OF BITS TO SUB-STREAMS FOR 8/9 CODE RATE WITH 128-QAM ......................109
TABLE 728 - PARAMETERS FOR DEMULTIPLEXING OF BITS TO SUB-STREAMS FOR 8/9 CODE RATE WITH 512-QAM ......................110
TABLE 729 - PARAMETERS FOR DEMULTIPLEXING OF BITS TO SUB-STREAMS FOR 8/9 CODE RATE WITH 2048-QAM ....................110
TABLE 730 - PARAMETERS FOR DEMULTIPLEXING OF BITS TO SUB-STREAMS FOR 8/9 CODE RATE WITH 8192-QAM ....................110
10/29/13
CableLabs
ix
CM-SP-PHYv3.1-I01-131029
TABLE 731 - PARAMETERS FOR DEMULTIPLEXING OF BITS TO SUB-STREAMS FOR 8/9 CODE RATE WITH 16384-QAM ..................110
TABLE 732 - SUBCARRIER SPACING ..................................................................................................................................123
TABLE 733 - K DEFINITIONS FOR 4K AND 8K FFT ...............................................................................................................123
TABLE 734 - CYCLIC PREFIX (CP) VALUES .........................................................................................................................125
TABLE 735 - ROLL-OFF PREFIX (RP) VALUES .....................................................................................................................125
TABLE 736 - RF OUTPUT ELECTRICAL REQUIREMENTS .........................................................................................................126
TABLE 737 - CMTS OUTPUT POWER...............................................................................................................................129
TABLE 738 - CMTS OUTPUT OUT-OF-BAND NOISE AND SPURIOUS EMISSIONS REQUIREMENTS .................................................132
TABLE 739 - CMTS OUTPUT REQUIREMENTS....................................................................................................................134
TABLE 740 - ELECTRICAL INPUT TO CM ............................................................................................................................136
TABLE 741 - CM MINIMUM CNR PERFORMANCE IN AWGN CHANNEL .................................................................................138
TABLE 742 - PLC COMPONENTS......................................................................................................................................139
TABLE 743 - PLC PREAMBLE FOR 4K FFT .........................................................................................................................142
TABLE 744 - PLC PREAMBLE FOR 8K FFT .........................................................................................................................143
TABLE 745 - SUBCARRIER DISTANCES FOR PLACEMENT OF PREDEFINED PILOTS ........................................................................162
TABLE 81 - MINISLOT PARAMETERS .................................................................................................................................167
TABLE 82 - DATA CODEWORD DEFINITION ........................................................................................................................173
TABLE 83 - NCP PARAMETERS........................................................................................................................................175
TABLE A1 - QAM CONSTELLATION SCALING FACTORS .........................................................................................................196
TABLE V1 - CMTS PROPOSED CONFIGURATION PARAMETERS ..............................................................................................208
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
1 SCOPE
1.1
Introduction and Purpose
This specification is part of the DOCSIS family of specifications developed by Cable Television Laboratories
(CableLabs). In particular, this specification is part of a series of specifications that defines the fourth generation of
high-speed data-over-cable systems, commonly referred to as the DOCSIS 3.1 specifications. This specification was
developed for the benefit of the cable industry, and includes contributions by operators and vendors from North and
South America, Europe and Asia.
This generation of the DOCSIS specifications builds upon the previous generations of DOCSIS specifications
(commonly referred to as the DOCSIS 3.0 and earlier specifications), leveraging the existing Media Access Control
(MAC) and Physical (PHY) layers, but with the addition of a new PHY layer designed to improve spectral
efficiency and provide better scaling for larger bandwidths (and appropriate updates to the MAC and management
layers to support the new PHY layer). It includes backward compatibility for the existing PHY layers in order to
enable a seamless migration to the new technology.
There are differences in the cable spectrum planning practices adopted for different networks in the world. For the
new PHY layer defined in this specification, there is flexibility to deploy the technology in any spectrum plan;
therefore, no special accommodation for different regions of the world is required for this new PHY layer.
However, due to the inclusion of the DOCSIS 3.0 PHY layers for backward compatibility purposes, there is still a
need for different region-specific physical layer technologies. Therefore, three options for physical layer
technologies are included in this specification, which have equal priority and are not required to be interoperable.
One technology option is based on the downstream channel identification plan that is deployed in North America
using 6 MHz spacing. The second technology option is based on the corresponding European multi-program
television distribution. The third technology option is based on the corresponding Chinese multi-program television
distribution. All three options have the same status, notwithstanding that the document structure does not reflect this
equal priority. The first of these options is defined in Sections 5 and 6, whereas the second is defined by replacing
the content of those sections with the content of Annex C. The third is defined by replacing the content of those
sections with the content of Annex D. Correspondingly, [ITU-T J.83-B] and [CEA-542-D] apply only to the first
option, and [EN 300 429] apply to the second and third. Compliance with this document requires compliance with
one of these implementations, but not with all three. It is not required that equipment built to one option shall
interoperate with equipment built to the other.
Compliance with frequency planning and EMC requirements is not covered by this specification and remains the
operators' responsibility. In this respect, [FCC15] and [FCC76] are relevant to the USA; [CAN/CSA CISPR 22-10]
and [ICES 003 Class A] to Canada; [EG 201 212], [EN 50083-1], [EN 50083-2], [EN 50083-7], [EN 61000-6-1],
and [EN 61000-6-3] are relevant to the European Union; [GB 8898-2011] and [GB/T 11318.1-1996] are relevant to
China.
1.2
Background
1.2.1
Broadband Access Network
A coaxial-based broadband access network is assumed. This may take the form of either an all-coax or hybridfiber/coax (HFC) network. The generic term "cable network" is used here to cover all cases.
A cable network uses a tree-and-branch architecture with analog transmission. The key functional characteristics
assumed in this document are the following:
Two-way transmission.
A maximum optical/electrical spacing between the CMTS and the most distant CM of 50 miles (80 km) in each
direction, although typical maximum separation may be 15 miles (24 km).
A maximum differential optical/electrical spacing between the closest and most distant modems of 50 miles (80
km) in each direction, although this would typically be limited to 15 miles (24 km).
10/29/13
CableLabs
CM-SP-PHYv3.1-I01-131029
At a propagation velocity in fiber of approximately 1.5 ns/ft. (5 ns/m), 50 miles (80 km) of fiber in each direction
results in a round-trip delay of approximately 0.8 ms. This is the maximum propagation delay assumed by this
specification.
1.2.2
1.2.2.1
Network and System Architecture

The DOCSIS Network
The elements that participate in the provisioning of DOCSIS services are shown in the following figure:
IPv4
CPE
NMS
CM
IPv6
CPE
HFC
CMTS
IPv4
CPE
CM
Provisioning
Systems
IPv6
CPE
Back Office Network
HFC Network
Home Network
Figure 11 - The DOCSIS Network
The CM connects to the operator's HFC network and to a home network, bridging packets between them. Many
CPEs can connect to the CMs' LAN interfaces. CPE can be embedded with the CM in a single device, or they can be
separated into standalone devices, as shown in Figure 11. CPE may use IPv4, IPv6 or both forms of IP addressing.
Examples of typical CPE are gateways, home routers, set-top devices, personal computers, etc.
The CMTS connects the operator's back office and core network to the HFC network. The CMTSs main function is
to forward packets between these two domains, and between upstream and downstream channels on the HFC
network.
Various applications are used to provide back office configuration and other support to the devices on the DOCSIS
network. These applications use IPv4 and/or IPv6 as appropriate to the particular operator's deployment. The
following applications include:
Provisioning Systems:
The DHCP servers provide the CM with initial configuration information, including the device IP address(es),
when the CM boots.
The Config File server is used to download configuration files to CMs when they boot. Configuration files are
in binary format and permit the configuration of the CM's parameters.
The Software Download server is used to download software upgrades to the CM.
The Time Protocol server provides Time Protocol clients, typically CMs, with the current time of day.
Certificate Revocation server provides certificate status.
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
NMS:
The SNMP Manager allows the operator to configure and monitor SNMP Agents, typically the CM and the
CMTS.
The Syslog server collects messages pertaining to the operation of devices.
The IPDR Collector server allows the operator to collect bulk statistics in an efficient manner.
1.2.3
Service Goals
As cable operators have widely deployed high-speed data services on cable television systems, the demand for
bandwidth has increased. To this end, CableLabs' member companies have decided to add new features to the
DOCSIS specification for the purpose of increasing capacity, increasing peak speeds, improving scalability,
enhancing network maintenance practices and deploying new service offerings.
The DOCSIS system allows transparent bi-directional transfer of Internet Protocol (IP) traffic, between the cable
system headend and customer locations, over an all-coaxial or HFC cable network. This is shown in simplified form
in Figure 12.
Figure 12 - Transparent IP Traffic Through the Data-Over-Cable System
1.2.4
Statement of Compatibility
This specification defines the DOCSIS 3.1 interface. Prior generations of DOCSIS were commonly referred to as the
DOCSIS 1.0, 1.1, 2.0, and 3.0 interfaces. DOCSIS 3.1 is backward-compatible with equipment built to the previous
specifications with the exception of DOCSIS 1.0 CMs. DOCSIS 3.1-compliant CMs interoperate seamlessly with
DOCSIS 3.0, 2.0, 1.1, and 1.0 CMTSs. DOCSIS 3.1-compliant CMTSs seamlessly support DOCSIS 3.0, DOCSIS
2.0, and DOCSIS 1.1 CMs.
10/29/13
CableLabs
CM-SP-PHYv3.1-I01-131029
1.2.5
Reference Architecture
Distribution Hub or Headend
DOCSIS Timing
Interface (DTI)
Edge
Resource
Manager
Edge Resource
Management
Interfaces
(ERMI)
Network Side
Interface (NSI)
M-CMTS
Core
Operations
Support System
DOCSIS
Timing Server
Operations Support
Systems Interface
(OSSI)
EQAM
Downstream
External-Phy
Interface (DEPI)
Upstream
Receiver
Downstream
RF Interface
(DRFI)
Cable Modem
to CPE
Interface
(CMCI)
M-CMTS
Downstream
RF
Network
Wide Area
Network
Fiber
Distribution
Rx
Opt.
Rx
Rx
Upstream
RF
Network
I-CMTS
Tx
Opt.
Tx
Tx
Fiber
Fiber
Node
Coax
Fiber
Node
Node Distribution
Physical Layer
Interface
(PHY)
NOTE: Lighter shaded areas are

related functionality, but out of the
scope of this document.
Cable
Modem
(CM)
Customer
Premises
Equipment
MAC & Upper Layer

Protocols Interface
(MULPI) &
Security Interface
(SEC)
Figure 1-3 - Data-Over-Cable Reference Architecture
The reference architecture for data-over-cable services and interfaces is shown in Figure 13.
1.2.6
DOCSIS 3.1 Documents
A list of the specifications in the DOCSIS 3.1 series is provided in Table 11. For further information, please refer
to http://www.cablemodem.com.
Table 11 - DOCSIS 3.1 Series of Specifications
Designation
Title
CM-SP-PHYv3.1
CM-SP-MULPIv3.1
Media Access Control and Upper Layer Protocols Interface Specification
CM-SP-OSSIv3.1-CM
Operations Support System Interface-CM Specification
CM-SP-OSSIv3.1-CMTS
Operations Support System Interface-CMTS Specification
CM-SP-SECv3.0
Security Specification
CM-SP-CMCIv3.0
Cable Modem CPE Interface Specification
This specification defines the interface for the physical layer.

Related DOCSIS specifications are listed in Table 12.
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Table 12 - DOCSIS 3.1 Related Specifications
Designation
Title
CM-SP-eDOCSIS
eDOCSIS Specification
CM-SP-DRFI
Downstream Radio Frequency Interface Specification
CM-SP-DTI
DOCSIS Timing Interface Specification
CM-SP-DEPI
Downstream External PHY Interface Specification
CM-SP-DSG
DOCSIS Set-Top Gateway Interface Specification
CM-SP-ERMI
Edge Resource Manager Interface Specification
CM-SP-M-OSSI
M-CMTS Operations Support System Interface Specification
CM-SP-L2VPN
Layer 2 Virtual Private Networks Specification
CM-SP-TEI
TDM Emulation Interfaces Specification
1.3
Requirements
Throughout this document, the words that are used to define the significance of particular requirements are
capitalized. These words are:
"MUST"
This word means that the item is an absolute requirement of this specification.
"MUST NOT"
This phrase means that the item is an absolute prohibition of this specification.
"SHOULD"
This word means that there may exist valid reasons in particular circumstances to ignore
this item, but the full implications should be understood and the case carefully weighed
before choosing a different course.
"SHOULD NOT"
This phrase means that there may exist valid reasons in particular circumstances when
the listed behavior is acceptable or even useful, but the full implications should be
understood and the case carefully weighed before implementing any behavior described
with this label.
"MAY"
This word means that this item is truly optional. One vendor may choose to include the
item because a particular marketplace requires it or because it enhances the product, for
example; another vendor may omit the same item.
This document defines many features and parameters, and a valid range for each parameter is usually specified.
Equipment (CM and CMTS) requirements are always explicitly stated. Equipment must comply with all mandatory
(MUST and MUST NOT) requirements to be considered compliant with this specification. Support of nonmandatory features and parameter values is optional.
1.4
Conventions
In this specification, the following convention applies any time a bit field is displayed in a figure. The bit field
should be interpreted by reading the figure from left to right, then top to bottom, with the most-significant bit (MSB)
being the first bit read, and the least-significant bit (LSB) being the last bit read.
10/29/13
CableLabs
CM-SP-PHYv3.1-I01-131029
1.5
Organization of Document
Section 1 provides an overview of the DOCSIS 3.1 series of specifications including the DOCSIS reference
architecture and statement of compatibility.
Section 2 includes a list of normative and informative references used within this specification.
Section 3 defines the terms used throughout this specification.
Section 4 defines the abbreviations and acronyms used throughout this specification.
Section 5 provides a technical overview and lists the key features of DOCSIS 3.1 technology for the functional area
of this specification; it also describes the key functional assumptions for the DOCSIS 3.1 system.
Section 6 defines the interface and related requirements for operation with legacy DOCSIS 3.0 (Single Carrier
QAM) channels only for each of: the CM downstream and upstream physical layer; and for the CMTS downstream
upstream physical layer.
Section 7 defines the interface and related requirements for operation with the new DOCSIS 3.1 channels, as well as
for combined operation of DOCSIS 3.0 and DOCSIS 3.1 channels. This is addressed for each of: the CM
downstream and upstream physical layer; and for the CMTS downstream upstream physical layer.
Section 8 defines PHY-MAC convergence how information is transferred between the MAC layer and the PHY
layer in both the upstream and downstream.
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
2 REFERENCES
2.1
Normative References
In order to claim compliance with this specification, it is necessary to conform to the following standards and other
works as indicated, in addition to the other requirements of this specification. Notwithstanding, intellectual property
rights may be required to use or implement such normative references.
All references are subject to revision, and parties to agreement based on this specification are encouraged to
investigate the possibility of applying the most recent editions of the documents listed below.
[CAN/CSA CISPR
22-10]
Information technology equipment - Radio disturbance characteristics - Limits and methods

of measurement (Adopted IEC CISPR 22:2008, sixth edition, 2008-09).
[CEA-542-D]
CEA-542-D: CEA Standard: Cable Television Channel Identification Plan, June 2013.
[DOCSIS DRFI]
Downstream Radio Frequency Interface Specification, CM-SP-DRFI-I13-130808, August

8, 2013, Cable Television Laboratories, Inc.
[DOCSIS
MULPIv3.1]
DOCSIS 3.1, MAC and Upper Layer Protocols Interface Specification, CM-SP-MULPIv3.1I01-131029, Cable Television Laboratories, Inc.
[DOCSIS PHYv3.0]
DOCSIS 3.0, Physical Layer Specification, CM-SP-PHYv3.0-I10-111117, November 11,

2011, Cable Television Laboratories, Inc.
[DVB-C2]
ETSI EN 302 769 V1.2.1: Digital Video Broadcasting (DVB); Frame structure channel
coding and modulation for a second generation digital transmission system for cable systems
(DVB-C2), April 2011.
[EG 201 212]
ETSI EG 201 212 V1.2.1: Electrical safety; Classification of interfaces for equipment to be
connected to telecommunication networks, November 1998.
[EN 300 429]
ETSI EN 300 429 V1.2.1: Digital Video Broadcasting (DVB); Framing structure, channel
coding and modulation for cable systems, April 1998.
[EN 50083-1]
CENELEC EN 50083-1: Cable networks for television signals, sound signals and interactive
services -- Part 1: Safety requirements, 2002.
[EN 50083-2]
services -- Part 2: Electromagnetic compatibility for equipment, 2005.
[EN 50083-7]
services -- Part 7: System performance, April 1996.
[EN 61000-6-1]
CENELEC EN 61000-6-4: Electromagnetic compatibility (EMC) -- Part 6-1: Generic

standards - Immunity for residential, commercial and light-industrial environments, October
2001.
[EN 61000-6-3]
CENELEC EN 61000-6-3: Electromagnetic compatibility (EMC) -- Part 6-3: Generic

standards - Emission standard for residential, commercial and light-industrial environments,
2003.
[FCC15]
Code of Federal Regulations, Title 47, Part 15, October 2005.
[FCC76]
Code of Federal Regulations, Title 47, Part 76, October 2005.
[GB 8898-2011]
Audio, video and similar electronic apparatusSafety requirements, Standardization

Administration of Peoples republic of China (SAC), www.sac.gov.cn
[ISO/IEC-61169-24]
ISO/IEC-61169-24, Radio-frequency connectors - Part 24: Sectional specification - Radio

frequency coaxial connectors with screw coupling, typically for use in 75 ohm cable
distribution systems (type F), 2001.
[ITU-T J.83-B]
Annex B to ITU-T Rec. J.83 (12/2007), Digital multi-program systems for television sound
and data services for cable distribution.
[SCTE 02]
ANSI/SCTE 02, Specification for "F" Port, Female Indoor, 2006.
10/29/13
CableLabs
CM-SP-PHYv3.1-I01-131029
2.2
Informative References
This specification uses the following informative references.

[CableLabs1]
Digital Transmission Characterization of Cable Television Systems, Cable Television

Laboratories, Inc., November 1994. URL:
http://www.cablelabs.com/downloads/digital_transmission.pdf
[CMB1993]
G. Castagnoli, S. Bruer, and M. Herrmann, "Optimization of Cyclic Redundancy-Check

Codes with 24 and 32 Parity Bits", IEEE Transactions on Communications, vol. 41, No. 6,
pp. 883-892, June 1993.
[PHYv3.1_CODECH CM-PHYv3.1_CODECHECK-131029, Number and size of codewords versus grant sizes,

ECK]
Cable Television Laboratories, Inc.
http://www.cablelabs.com/cablemodem/specifications/specifications31.html
[DOCSIS OSSIv3.1] DOCSIS 3.1 Operations Support System Interface-CM Specification, forthcoming, Cable
Television Laboratories, Inc. and DOCSIS 3.1 Operations Support System Interface-CMTS
Specification, forthcoming, Cable Television Laboratories, Inc.
[GB/T 11318.11996]
Equipments and components used in cabled distribution systems primarily intended for
television and sound signals--Part 1:Generic specifications, China Zhijian Publish House
SAC.
[ICES 003 Class A]
Information Technology Equipment (ITE) Limits and methods of measurement.
[SCTE]
SCTE Measurement Recommended Practices for Cable Systems, Fourth Edition, March
2012.
2.3
Reference Acquisition
American National Standards Institute (ANSI), http://www.ansi.org
Cable Television Laboratories, Inc., 858 Coal Creek Circle, Louisville, CO 80027;
Phone +1-303-661-9100; Fax +1-303-661-9199; http://www.cablelabs.com
CENELEC: European Committee for Electro-technical Standardization, http://www.cenelec.org
Ecma International, http://www.ecma-international.org/
EIA: Electronic Industries Alliance, http://www.eia.org/new_contact/
ETSI: European Telecommunications Standards Institute,

http://www.etsi.org/services_products/freestandard/home.htm
Internet Engineering Task Force (IETF) Secretariat, 48377 Fremont Blvd., Suite 117, Fremont, California
94538, USA, Phone: +1-510-492-4080, Fax: +1-510-492-4001, http://www.ietf.org
ISO: International Organization for Standardization (ISO), http://www.iso.org/iso/en/xsite/contact/contact.html
ITU: International Telecommunications Union (ITU), http://www.itu.int/home/contact/index.html
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
3 TERMS AND DEFINITIONS

This specification uses the following terms:
Active Subcarrier
Any subcarrier other than an excluded subcarrier.
Adaptive Equalizer
A circuit in a digital receiver that compensates for channel response

impairments. In effect, the circuit creates a digital filter that has approximately
the opposite complex frequency response of the channel through which the
desired signal was transmitted.
Adaptive Equalizer Tap
See tap.
Adaptive Pre-Equalizer
A circuit in a DOCSIS 1.1 or newer cable modem that pre-equalizes or predistorts the transmitted upstream signal to compensate for channel response
impairments. In effect, the circuit creates a digital filter that has approximately
the opposite complex frequency response of the channel through which the
desired signal is to be transmitted.
Additive White Gaussian

Noise (AWGN)
See thermal noise.
Availability
The ratio of time that a service, device, or network is available for use to total
time, usually expressed as a percentage of the total time. For example, fournines availability, written as 99.99%, means that a service is available 8759.12
hours out of 8760 total hours in a year.
BCH
A class of error correction codes named after the inventors Raj Bose, D. K.
Ray-Chaudhuri, and Alexis Hocquenghem.
Binary Phase Shift Keying

(BPSK)
A form of digital modulation in which two phases separated by 180 degrees

support the transmission of one bit per symbol.
Bit Error Rate (BER)
See bit error ratio.
Bit Error Ratio (BER)
The ratio of errored bits to the total number of bits transmitted, received, or
processed over a defined amount of time. Mathematically, BER = (number of
errored bits)/(total number of bits) or BER = (error count in measurement
period)/(bit rate * measurement period). Usually expressed in scientific
notation format. Also called bit error rate.
Bit Loading
The technique of assigning the optimum number of bits (modulation order) for
transmission per subcarrier in an OFDM or OFDMA system.
Burst
A single continuous RF signal from the cable modem upstream transmitter,

from transmitter on to transmitter off.
Burst Noise
1) Another name for impulse noise. 2) A type of noise comprising random and
sudden step-like changes between levels, often occurring in semiconductors.
Sometimes called popcorn noise.
Cable Modem (CM)
A modulator-demodulator at the subscriber premises intended for use in

conveying data communications on a cable television system.
Cable Modem Termination

System (CMTS)
A device located at the cable television system headend or distribution hub,

which provides complementary functionality to the cable modems to enable
data connectivity to a wide-area network.
Carrier-To-Noise Ratio
(CNR or C/N)
The ratio of signal (or carrier) power to noise power in a defined measurement
bandwidth. For digital modulation, CNR = ES/N0, the energy per symbol-tonoise power spectral density ratio; the signal (or carrier) power is measured in
the occupied bandwidth, and the noise power is normalized to the modulationrate bandwidth. For NTSC analog video, the visual carrier power measurement
bandwidth is nominally 300 kHz, and the noise power measurement bandwidth
is 4 MHz.
10/29/13
CableLabs
CM-SP-PHYv3.1-I01-131029
CEA-542-D
A Consumer Electronics Association standard that defines a channel

identification plan for 6 MHz-wide channel frequency allocations in cable
systems.
Ceiling (Ceil)
A mathematical function that returns the lowest-valued integer that is greater

than or equal to a given value.
Channel
A portion of the electromagnetic spectrum used to convey one or more RF

signals between a transmitter and receiver. May be specified by parameters
such as center frequency, bandwidth, or CEA channel number.
Codeword
Forward error correction data block, comprising a combination of information

bytes and parity bytes.
Codeword Error Ratio

(CER)
The ratio of errored codewords to the total number of codewords transmitted,

received, or processed over a defined amount of time. Mathematically, CER =
(number of errored codewords)/(total number of codewords). Usually
expressed in scientific notation format.
Coefficient
Complex number that establishes the gain of each tap in an adaptive equalizer
or adaptive pre-equalizer.
Common Path Distortion

(CPD)
Clusters of second and third order distortion beats generated in a diode-like

nonlinearity such as a corroded connector in the signal path common to
downstream and upstream. The beats tend to be prevalent in the upstream
spectrum. When the primary RF signals are digitally modulated signals instead
of analog TV channels, the distortions are noise-like rather than clusters of
discrete beats.
Complementary Pilots
Subcarriers that carry data, but with a lower modulation order than other data
subcarriers in a given minislot. Complementary pilots allow phase tracking
along the time axis for frequency offset and phase noise correction, and may be
used by the CMTS upstream receiver to enhance signal processing, such as
improving the accuracy of center frequency offset acquisition.
Composite Second Order

(CSO)
Clusters of second order distortion beats generated in cable network active

devices that carry multiple RF signals. When the primary RF signals are
digitally modulated signals instead of analog TV channels, the distortions are
noise-like rather than clusters of discrete beats.
Composite Triple Beat

(CTB)
Clusters of third order distortion beats generated in cable network active

devices that carry multiple RF signals. When the primary RF signals are
digitally modulated signals instead of analog TV channels, the distortions are
noise-like rather than clusters of discrete beats.
Continuous Pilots
Pilots that occur at the same frequency location in every OFDM symbol, and
which are used for frequency and phase synchronization.
Convolution
A process of combining two signals in which one of the signals is timereversed and correlated with the other signal. The output of a filter is the
convolution of its impulse response with the input signal.
Convolutional Interleaver
An interleaver in which symbols are sequentially shifted into a bank of "I"

registers. Each successive register has "J" symbols more storage than the
preceding register. The first interleaver path has zero delay, the second has a J
symbol period of delay, the third 2 x J symbol periods of delay, etc., up to the
Ith path which has (I - 1) x J symbol periods of delay. This process is reversed
in the receivers deinterleaver so that the net delay of each symbol is the same
through the interleaver and deinterleaver. See also interleaver.
Correlation
1) A process of combining two signals in which the signals are multiplied

sample-by-sample and summed; the process is repeated at each sample as one
signal is slid in time past the other. 2) Cross-correlation is a measure of
similarity between two signals.
10
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Cross Modulation (XMOD)
A form of television signal distortion in which modulation from one or more

television signals is imposed on another signal or signals.
Customer Premises
Equipment (CPE)
Device such as a cable modem or set-top at the subscribers or other end users
location. May be provided by the end user or the service provider.
Cyclic Prefix (CP)
A copy of the end of a symbol that is added to the beginning of the same
symbol, in order to help mitigate the effects of micro-reflections and similar
impairments.
Decibel (dB)
Ratio of two power levels expressed mathematically as dB = 10log10(P1/P2).
Decibel Carrier (dBc)
Ratio of the power of a signal to the power of a reference carrier, expressed

mathematically as dBc = 10log10(Psignal/Pcarrier).
Decibel Millivolt (dBmV)
Unit of RF power expressed in terms of voltage, defined as decibels relative to

1 millivolt, where 1 millivolt equals 13.33 nanowatts in a 75 ohm impedance.
Mathematically, dBmV = 20log10(value in mV/1 mV).
Decibel Reference (dBr)
Ratio of a signal level to a reference signal level. When the signals are noise or
noise-like, the measurement bandwidth for the two signals is the same. When
both signal levels are in the same units of power, the ratio is expressed
mathematically as dBr = 10log10(Psignal/Preference). When both signal levels are in
the same units of voltage, and assuming the same impedance, the ratio is
expressed mathematically as dBr = 20log10(Vsignal/Vreference).
Discrete Fourier Transform

(DFT)
Part of the family of mathematical methods known as Fourier analysis, which

defines the "decomposition" of signals into sinusoids. Discrete Fourier
transform defines the transformation from the time to the frequency domain.
See also inverse discrete Fourier transform.
Distortion
See linear distortion and nonlinear distortion.
Distribution Hub
A facility in a cable network which performs the functions of a headend for

customers in its immediate area, and which receives some or all of its content
for transmission from a master headend in the same metropolitan or regional
area.
DOCSIS
Data-Over-Cable Service Interface Specifications. A group of specifications

that defines interoperability between cable modem termination systems and
cable modems.
DOCSIS 1.x
Abbreviation for DOCSIS versions 1.0 or 1.1.
Downstream
1) The direction of RF signal transmission from headend or hub site to

subscriber. In North American cable networks, the downstream or forward
spectrum may occupy frequencies from just below 54 MHz to as high as 1002
MHz or more. 2) The DOCSIS 3.1 downstream is 258 MHz (optional 108
MHz) to 1218 MHz (optional 1794 MHz).
Downstream Channel
A portion of the electromagnetic spectrum used to convey one or more RF

signals from the headend or hub site to the subscriber premises. For example, a
single CEA channels bandwidth is 6 MHz, and a downstream DOCSIS 3.1
OFDM channels bandwidth can be up to 192 MHz.
Drop
Coaxial cable and related hardware that connects a residence or other service
location to a tap in the nearest coaxial feeder cable. Also called drop cable or
subscriber drop.
Dynamic Host
Configuration Protocol
(DHCP)
A protocol that defines the dynamic or temporary assignment of Internet

protocol addresses, so that the addresses may be reused when they are no
longer needed by the devices to which they were originally assigned.
Dynamic Range Window

(DRW)
The range, in decibels, defining the maximum power difference between

multiple transmitters in a cable modem that is operating in multiple transmit
channel mode.
10/29/13
CableLabs
11
CM-SP-PHYv3.1-I01-131029
Encompassed Spectrum
RF bandwidth equal to 204.8 MHz minus the number of subcarriers in the

band-edge exclusion subband for the upper and lower band edges combined,
multiplied by the subcarrier spacing in MHz (e.g., 25 kHz spacing = 0.025
MHz). Encompassed spectrum also is equal to the center frequency of the
OFDM channels highest frequency subcarrier that is modulated, minus the
center frequency of the OFDM channels lowest frequency subcarrier that is
modulated, plus the subcarrier spacing.
Equalizer Tap
See tap.
Equivalent Legacy DOCSIS

Channels
Within a downstream OFDM channel, an integer number equal to

ceil(modulated spectrum in MHz / 6).
Excluded Subcarrier
Subcarrier that cannot be used because another type of service is using the
subcarriers frequency or a permanent ingressor is present on the frequency.
The CMTS or cable modem is administratively configured to not transmit on
excluded subcarriers.
Exclusion Band
A set of contiguous subcarriers within the OFDM or OFDMA channel

bandwidth that are set to zero-value by the transmitter to reduce interference to
other co-existing transmissions such as legacy SC-QAM signals.
F Connector
A threaded, nominally 75-ohm impedance RF connector, whose electrical and

physical specifications are defined in various SCTE standards. Commonly used
on smaller sizes of coaxial cable such as 59-series and 6-series, and on mating
interfaces of subscriber drop components, customer premises equipment, and
some headend and test equipment.
Fast Fourier Transform

(FFT)
An algorithm to compute the discrete Fourier transform from the time domain
to the frequency domain, typically far more efficiently than methods such as
correlation or solving simultaneous linear equations. See also discrete Fourier
transform, inverse discrete Fourier transform, and inverse fast Fourier
transform.
FFT Duration
Reciprocal of subcarrier spacing. For example, with 50 kHz subcarrier spacing,

FFT duration is 20 s, and with 25 kHz subcarrier spacing, FFT duration is 40
s. Sometimes called "useful symbol duration." See also symbol duration.
Fiber Node
See node.
Filler Subcarrier
A zero bit loaded subcarrier that is inserted in an OFDM symbol when

no data is transmitted, or when the number of codewords has exceeded
the upper limit, or when it is not possible to begin a new codeword
because of insufficient space to include a next codeword pointer.
Floor
A mathematical function that returns the highest-valued integer that is less than
or equal to a given value.
Forward
See downstream.
Forward Error Correction

(FEC)
A method of error detection and correction in which redundant information is

sent with a data payload in order to allow the receiver to reconstruct the
original data if an error occurs during transmission.
Frame
1) A logical grouping of information sent as an Open System Interconnection

Layer 2 unit over a transmission medium. 2) In OFDM transmission, definition.
3) In OFDMA transmission, a logical grouping of upstream data and pilots
comprising a configurable number of symbols (minimum of 6 to a maximum of
32) and minislots (minimum of 4 subcarriers per minislot to a maximum of 16
subcarriers per minislot). Multiple modems can share the same OFDMA frame
simultaneously by transmitting data and pilots on allocated subcarriers within
the frame.
12
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Frequency Division
Multiple Access (FDMA)
A multiple access technology that accommodates multiple users by allocating

each users traffic to one or more discrete frequency bands, channels, or
subcarriers.
Frequency Division
Multiplexing (FDM)
The transmission of multiple signals through the same medium at the same
time. Each signal is on a separate frequency or assigned to its own channel. For
example, an analog TV signal might be carried on CEA channel 7 (174 MHz180 MHz), a 256-QAM digital video signal on channel 8 (180-186 MHz), and
so on.
Frequency Response
A complex quantity describing the flatness of a channel or specified frequency

range, and which has two components: amplitude (magnitude)-versusfrequency, and phase-versus-frequency.
Gigahertz (GHz)
One billion (109) hertz. See also hertz.
Group Delay (GD)
The negative derivative of phase with respect to frequency, expressed

mathematically as GD = -(d/d) and stated in units of time such as
nanoseconds or microseconds.
Group Delay Ripple
Group delay variation which has a sinusoidal or scalloped sinusoidal shape

across a specified frequency range.
Group Delay Variation

(GDV) or Group Delay
Distortion
The difference in propagation time between one frequency and another in a

circuit, device, or system. That is, some frequency components of a signal may
arrive at the output of the circuit, device, or system before other frequency
components, causing distortion of the signal.
Guard Interval
In the time domain, the period from the end of one symbol to the beginning of
the next symbol, which includes the cyclic prefix and applied transmit
windowing. Also called guard time.
Guard Band
A narrow range of frequencies in which user data is not transmitted, located at

the lower and upper edges of a channel, at the lower and upper edges of a gap
within a channel, or in between channels, in order to minimize interference.
Harmonic Related Carriers

(HRC)
A method of spacing channels on a cable television system defined in [CEA542-D], in which visual carriers are multiples of 6.0003 MHz. A variation of
HRC channelization used in some European cable networks is based upon
multiples of 8 MHz.
Headend
A central facility that is used for receiving, processing, and combining

broadcast, narrowcast and other signals to be carried on a cable network.
Somewhat analogous to a telephone companys central office. Location from
which the DOCSIS cable plant fans out to subscribers. See also distribution
hub.
Header
Protocol control information located at the beginning of a protocol data unit.
Hertz (Hz)
A unit of frequency equivalent to one cycle per second.
Hum Modulation
Amplitude distortion of a signal caused by the modulation of that signal by

components of the power source (e.g., 60 Hz) and/or its harmonics.
Hybrid Fiber/Coax (HFC)
A broadband bidirectional shared-media transmission system or network

architecture using optical fibers between the headend and fiber nodes, and
coaxial cable distribution from the fiber nodes to the subscriber locations.
Impedance
The combined opposition to current in a component, circuit, device, or

transmission line that contains both resistance and reactance. Represented by
the symbol Z and expressed in ohms.
10/29/13
CableLabs
13
CM-SP-PHYv3.1-I01-131029
Impulse Noise
Noise that is bursty in nature, characterized by non-overlapping transient

disturbances. May be periodic (e.g., automobile ignition noise or high-voltage
power line corona noise), or random (e.g., switching noise or atmospheric noise
from thunderstorms). Generally of short duration from about 1 microsecond
to a few tens of microseconds with a fast risetime and moderately fast
falltime.
Incremental Related
Carriers (IRC)
A method of spacing channels on a cable television system defined in [CEA542-D], in which all visual carriers except channels 5 and 6 are offset +12.5
kHz with respect to the standard channel plan. Channels 5 and 6 are offset
+2.0125 MHz with respect to the standard channel plan. See also standard
frequencies.
In-Phase (I)
The real part of a vector that represents a signal, with 0 degrees phase angle
relative to a reference carrier. See also quadrature (Q).
Interleaver
A subset or layer of the forward error correction process, in which the data to
be transmitted is rearranged or mixed such that the original bits, bytes, or
symbols are no longer adjacent. The latter provides improved resistance to
various forms of interference, especially burst or impulse noise. Interleaving
may be performed in the time domain, frequency domain, or both. A deinterleaver in the receiver rearranges the bits, bytes, or symbols into their
original order prior to additional error correction.
International
Electrotechnical
Commission (IEC)
An organization that prepares and publishes international standards for

electrical, electronic and related technologies.
International Organization
for Standardization (ISO)
An organization that develops voluntary international standards for technology,

business, manufacturing, and other industries.
Internet Engineering Task

Force (IETF)
A body responsible, among other things, for developing standards used in the
Internet.
Internet Protocol (IP)
A network layer protocol that supports connectionless internetwork service,

and which contains addressing and control information that allows packets to
be routed. Widely used in the public Internet as well as private networks. The
vast majority of IP devices support IP version 4 (IPv4) defined in RFC-791,
although support for IP version 6 (IPv6, RFC-2460) continues to increase.
Internet Protocol Detail

Record
The record formatter and exporter functions of the CMTS that provides
information about Internet protocol-based service usage, and other activities
that can be used by operational support systems and business support systems.
Inverse Discrete Fourier

Transform (IDFT)
Part of the family of mathematical methods known as Fourier analysis, which

defines the "decomposition" of signals into sinusoids. Inverse discrete Fourier
transform defines the transformation from the frequency to the time domain.
See also discrete Fourier transform.
Inverse Fast Fourier

Transform (IFFT)
An algorithm to compute the inverse discrete Fourier transform from the

frequency domain to the time domain, typically far more efficiently than
methods such as correlation or solving simultaneous linear equations. See also
discrete Fourier transform, fast Fourier transform, and inverse discrete
Fourier transform.
Jitter
The fluctuation in the arrival time of a regularly scheduled event such as a

clock edge or a packet in a stream of packets. Jitter is defined as fluctuations
above 10 Hz.
Kilohertz (kHz)
One thousand (103) hertz. See also hertz.
14
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Latency
1) The time taken for a signal element to propagate through a transmission

medium or device. 2) The delay between a devices request for network access
and when permission is granted for transmission. 3) The delay from when a
frame is received by a device to when the frame is forwarded via the devices
destination port.
Layer
One of seven subdivisions of the Open System Interconnection reference

model.
Linear Distortion
A class of distortions that occurs when the overall response of the system
(including transmitter, cable plant, and receiver) differs from the ideal or
desired response. This class of distortions maintains a linear, or 1:1, signal-todistortion relationship (increasing signal by 1 dB causes distortion to increase
by 1 dB), and often occurs when amplitude-versus-frequency and/or phaseversus-frequency depart from ideal. Linear distortions include impairments
such as micro-reflections, amplitude ripple, and group delay variation, and can
be corrected by an adaptive equalizer.
Low Density Parity Check

(LDPC)
An error correction code used in DOCSIS 3.1. LDPC is more robust than ReedSolomon error correction codes.
MAC Address
The "built-in" hardware address of a device connected to a shared medium.
MAC Frame
MAC header plus optional protocol data unit.
MAC Management
Message (MMM)
Unclassified traffic between the CMTS and cable modem. Examples include
MAC domain descriptor, ranging-request, ranging-response, and upstream
channel descriptor messages.
Media Access Control

(MAC)
A sublayer of the Open Systems Interconnection models data link layer (Layer
2), which manages access to shared media such as the Open Systems
Interconnection models physical layer (Layer 1).
Megahertz (MHz)
One million (106) hertz. See also hertz.
Micro-reflection
A short time delay echo or reflection caused by an impedance mismatch. A

micro-reflections time delay is typically in the range from less than a symbol
period to several symbol periods.
Microsecond (s)
One millionth (10-6) of a second
Millisecond (ms)
One thousandth (10-3) of a second
Millivolt (mV)
One thousandth (10-3) of a volt
Minislot
In DOCSIS 3.0 and earlier TDMA applications, a unit of time for upstream
transmission that is an integer multiple of 6.25 s units of time called "ticks."
In DOCSIS 3.1 OFDMA applications, a group of dedicated subcarriers, all with
the same modulation order, for upstream transmission by a given cable modem.
For both TDMA and OFDMA, a cable modem may be assigned one or more
minislots in a transmission burst.
Modulated Spectrum
Encompassed spectrum minus the total spectrum in the internal excluded

subbands of the OFDM channel, where the total spectrum in the internal
excluded subbands is equal to the number of subcarriers in all of the internal
excluded subbands of the OFDM channel multiplied by the subcarrier spacing.
10/29/13
CableLabs
15
CM-SP-PHYv3.1-I01-131029
Modulation Error Ratio

(MER)
The ratio of average signal constellation power to average constellation error

power that is, digital complex baseband signal-to-noise ratio expressed in
decibels. In effect, MER is a measure of how spread out the symbol points in a
constellation are. More specifically, MER is a measure of the cluster variance
that exists in a transmitted or received waveform at the output of an ideal
receive matched filter. MER includes the effects of all discrete spurious, noise,
carrier leakage, clock lines, synthesizer products, linear and nonlinear
distortions, other undesired transmitter and receiver products, ingress, and
similar in-channel impairments.
Modulation Rate
The signaling rate of the upstream modulator (for example, 1280 to 5120 kHz).
In S-CDMA it is the chip rate. In TDMA, it is the channel symbol rate.
Multiple Transmit Channel

(MTC) [Mode]
Operational mode in a cable modem that enables the simultaneous transmission

of more than one upstream channel.
Nanosecond (ns)
One billionth (10-9) of a second
National Television System

Committee (NTSC)
The committee that defined the analog television broadcast standards (black
and white in 1941, color in 1953) used in North America and some other parts
of the world. The NTSC standards are named after the committee.
Next Codeword Pointer

(NCP)
A message block used to identify where a codeword begins.
Node
An optical-to-electrical (RF) interface between a fiber optic cable and the

coaxial cable distribution network. Also called fiber node.
Noise
Typically any undesired signal or signalsother than discrete carriers or

discrete distortion productsin a device, communications circuit, channel or
other specified frequency range. See also impulse noise, phase noise, and
thermal noise.
Nonlinear Distortion
A class of distortions caused by a combination of small signal nonlinearities in

active devices and by signal compression that occurs as RF output levels reach
the active devices saturation point. Nonlinear distortions generally have a
nonlinear signal-to-distortion amplitude relationshipfor instance, 1:2, 1:3 or
worse (increasing signal level by 1 dB causes distortion to increase by 2 dB, 3
dB, or more). The most common nonlinear distortions are even order
distortions such as composite second order, and odd order distortions such as
composite triple beat. Passive components can generate nonlinear distortions
under certain circumstances.
Occupied Bandwidth
Total spectrum (including guard bands, roll-off regions, and unoccupied

exclusion bands) used for DOCSIS 3.1 RF signaling and not being utilized by
other technologies.
Orthogonal
Distinguishable from or independent such that there is no interaction or

interference. In OFDM, subcarrier orthogonality is achieved by spacing the
subcarriers at the reciprocal of the symbol period (T), also called symbol
duration time. This spacing results in the sinc (sin x/x) frequency response
curves of the subcarriers lining up so that the peak of one subcarriers response
curve falls on the first nulls of the lower and upper adjacent subcarriers
response curves. Orthogonal subcarriers each have exactly an integer number
of cycles in the interval T.
Orthogonal Frequency
Division Multiple Access
(OFDMA)
An OFDM-based multiple-access scheme in which different subcarriers or

groups of subcarriers are assigned to different users.
16
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Orthogonal Frequency
Division Multiplexing
(OFDM)
A data transmission method in which a large number of closely-spaced or

overlapping very-narrow-bandwidth orthogonal QAM signals are transmitted
within a given channel. Each of the QAM signals, called a subcarrier, carries a
small percentage of the total payload at a very low data rate.
Phase Noise
Rapid, short-term, random fluctuations in the phase of a wave, caused by time

domain instabilities.
PHY Link Channel (PLC)
A set of contiguous OFDM subcarriers (eight for 4K FFT and 16 for 8K FFT),
constituting a "sub-channel" of the OFDM channel, which conveys physical
layer parameters from the CMTS to cable modem.
Physical Layer (PHY)
Layer 1 in the Open System Interconnection architecture; the layer that

provides services to transmit bits or groups of bits over a transmission link
between open systems and which entails electrical, mechanical and
handshaking procedures.
Picosecond (ps)
One trillionth (10-12) of a second
Pilot
A dedicated OFDM subcarrier that may be used for such purposes as channel
estimation (measurement of channel condition), synchronization, and other
purposes. See also complementary pilots, continuous pilots and scattered pilots.
Pload_min_set
Reduction of channel n maximum transmit power with respect to Phi_n, when

multiple transmit channel mode is enabled. The value of Pload_min_set is sent
by the CMTS to the cable modem.
Pload_n
Reduction of channel n reported transmit power with respect to Phi_n, where

Phi_n = min(Pmax - Gconst) over all burst profiles used by the cable modem in
channel n.
Preamble
A data sequence transmitted at or near the beginning of a frame, allowing the

receiver time to achieve lock and synchronization of transmit and receive
clocks.
Pre-equalizer
See adaptive pre-equalizer.
profile
A set of parameters that defines how information is transmitted from a CMTS

to a cable modem, or from a cable modem to a CMTS. Examples of some of
the parameters defined in a profile include modulation order, forward error
correction, preamble, and guard interval.
Protocol
A description of a set of rules and formats that specify how devices on a

network exchange data.
Pseudo Random Binary

Sequence (PRBS)
A deterministic sequence of bits that appears to be random, that is, with no

apparent pattern. Also called pseudo random bit stream.
QAM Signal
Analog RF signal that uses quadrature amplitude modulation to convey

information such as digital data.
Quadrature (Q)
The imaginary part of a vector that represents a signal, with 90 degrees phase
angle relative to a reference carrier. See also in-phase (I).
Quadrature Amplitude
Modulation (QAM)
A modulation technique in which an analog signals amplitude and phase vary

to convey information, such as digital data. The name "quadrature" indicates
that amplitude and phase can be represented in rectangular coordinates as inphase (I) and quadrature (Q) components of a signal.
Quadrature Phase Shift

Keying (QPSK)
A form of digital modulation in which four phase states separated by 90

degrees support the transmission of two bits per symbol. Also called 4-QAM.
Radio Frequency (RF)
That portion of the electromagnetic spectrum from a few kilohertz to just below
the frequency of infrared light.
10/29/13
CableLabs
17
CM-SP-PHYv3.1-I01-131029
Randomizer
A subset or layer of the forward error correction process, in which the data to
be transmitted is randomized using a PRBS scrambler. Randomization spreads
out the energy across the spectrum, ensures uniform population of all of the
data constellation points, and minimizes the likelihood of long strings of all
zeros or ones.
Reed-Solomon (R-S)
A class of error correction codes named after the inventors Irving Reed and
Gustave Solomon. The forward error correction in DOCSIS 3.0 and earlier uses
Reed-Solomon error correction codes.
Return
See upstream.
Return Loss (R)
The ratio of incident power PI to reflected power PR, expressed mathematically

as R = 10log10(PI/PR), where R is return loss in decibels.
Reverse
See upstream.
RF Channel
See channel.
Root Mean Square (RMS)
A statistical measure of the magnitude of a varying quantity such as current or

voltage, where the RMS value of a set of instantaneous values over, say, one
cycle of alternating current is equal to the square root of the mean value of the
squares of the original values.
Scattered Pilots
Pilots that do not occur at the same frequency in every symbol, and which may
be used for channel estimation. The locations of scattered pilots change from
one OFDM symbol to another.
Scrambler
See randomizer.
Signal-To-Composite Noise
(SNC)
The ratio of signal power to composite noise power in a defined measurement

bandwidth, where composite noise is the combination of thermal noise and
composite intermodulation distortion (noise-like distortion).
Single Carrier Quadrature

Amplitude Modulation
(SC-QAM)
Data transmission method used in DOCSIS 1.x, 2.0 and 3.0, in which each
downstream or upstream RF channel slot carries only one QAM signal.
Society of Cable
Telecommunications
Engineers (SCTE)
A non-profit professional association that specializes in professional

development, standards, certification, and information for the cable
telecommunications industry.
Standard Frequencies
(STD)
Method of spacing channels on a cable television system defined in [CEA-542D]. Channels 2-6 and 7-13 use the same frequencies as over-the-air channels 26 and 7-13. Other cable channels below Ch. 7 down to 91.25 MHz and above
Ch. 13 are spaced in 6 MHz increments.
Subcarrier
One of a large number of closely spaced or overlapping orthogonal narrowbandwidth data signals within an OFDM channel. Also called a tone. See also
excluded subcarrier, unused subcarrier, and used subcarrier.
Sublayer
A subdivision of a layer in the Open System Interconnection reference model.
Subscriber
End user or customer connected to a cable network.
Subslot
A subdivision or subunit of a minislot that fits within a minislots boundaries,

used to provide multiple transmission opportunities for bandwidth requests.
Data subcarriers within a subslot use QPSK, and are not FEC encoded.
Symbol Duration
Sum of one subcarriers clock period and cyclic prefix duration. Symbol
duration is greater than FFT duration, because symbol duration includes a
prepended cyclic prefix.
Synchronous Code Division

Multiple Access (S-CDMA)
A multiple access physical layer technology in which different transmitters can

share a channel simultaneously. The individual transmissions are kept distinct
by assigning each transmission an orthogonal "code." Orthogonality is
maintained by all transmitters being precisely synchronized with one another.
18
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Tap
(1) In the feeder portion of a coaxial cable distribution network, a passive

device that comprises a combination of a directional coupler and splitter to
"tap" off some of the feeder cable RF signal for connection to the subscriber
drop. So-called self-terminating taps used at feeder ends-of-line are splitters
only and do not usually contain a directional coupler. Also called a multitap.
(2) The part of an adaptive equalizer where some of the main signal is "tapped"
off, and which includes a delay element and multiplier. The gain of the
multipliers is set by the equalizers coefficients. (3) One term of the difference
equation in a finite impulse response or an infinite impulse response filter. The
difference equation of a FIR follows: y(n) = b0x(n) + b1x(n-1) + b2x(n-2) + +
bNx(n-N).
Thermal Noise
The fluctuating voltage across a resistance due to the random motion of free
charge caused by thermal agitation. Also called Johnson-Nyquist noise. When
the probability distribution of the voltage is Gaussian, the noise is called
additive white Gaussian noise (AWGN).
Time Division Multiple

Access (TDMA)
A multiple access technology that enables multiple users to access, in sequence,

a single RF channel by allocating unique time slots to each user of the channel.
Transit Delay
The time required for a signal to propagate or travel from one point in a
network to another point in the network, for example, from the CMTS to the
most distant cable modem. Also called propagation delay.
Transmit Channel Set

(TCS)
The combination of legacy SC-QAM channels and OFDMA channels that may
be transmitted by a cable modem.
Transmit Pre-Equalizer
See adaptive pre-equalizer.
Under-Grant Hold
Bandwidth (UGHB)
The minimum grant bandwidth that can be allocated beyond which the spurious
emissions limits (in dBc) are no longer relaxed as the based on grant size
continues to decrease. Defined mathematically as UGHB = (100% grant
spectrum)/(under-grant hold number of users).
Under-Grant Hold Number

of Users (UGHU)
The maximum number of equal-size grants that can be allocated beyond which
the spurious emissions limits (in dBc) are no longer relaxed as the number of
based on grants size continues to increase. Defined mathematically as UGHU =
floor[0.2 + 10((-44 - SpurFloor)/10)]
Unused Subcarrier
An upstream subcarrier that could be used at a future time for data traffic but is
not currently part of a minislot. The cable modem transmits only probes on
these subcarriers and only when instructed to do so by MAP messages. MULPI
term
Upstream
1) The direction of RF signal transmission from subscriber to headend or hub

site. Also called return or reverse. In most North American cable networks, the
legacy upstream spectrum occupies frequencies from 5 MHz to as high as 42
MHz. 2) The DOCSIS 3.1 upstream is 5-204 MHz, with support for 5-42 MHz,
5-65 MHz, 5-85 MHz and 5-117 MHz.
Upstream Channel
1) The direction of RF signal transmission from subscriber to headend or hub

site. Also called return or reverse. In most North American cable networks, the
legacy upstream spectrum occupies frequencies from 5 MHz to as high as 42
MHz. 2) The DOCSIS 3.1 upstream is 5-204 MHz, with support for 5-42 MHz,
5-65 MHz, 5-85 MHz and 5-117 MHz.
Upstream Channel
Descriptor (UCD)
The MAC management message used to communicate the characteristics of the

upstream physical layer to the cable modems.
Used Subcarrier
An upstream subcarrier that is part of a minislot. The cable modem transmits

data, ranging, and probes on these subcarriers when instructed to do so by MAP
messages. MULPI term
Useful Symbol Duration
See FFT duration.
10/29/13
CableLabs
19
CM-SP-PHYv3.1-I01-131029
Vector
A quantity that expresses magnitude and direction (or phase), and is

represented graphically using an arrow.
Windowing
A technique to shape data in the time domain, in which a segment of the start of
the IFFT output is appended to the end of the IFFT output to taper or roll-off
the edges of the data using a raised cosine function. Windowing maximizes the
capacity of the channel by sharpening the edges of the OFDM/A signal in the
frequency domain.
Word
Information part of a codeword, without parity. See also codeword.
Zero Bit Loaded Minislot
A minislot that does not carry user data, but does carry pilots.
Zero Bit-LoadedSubcarrier
A subcarrier with power but not carrying user data, although it could be
modulated by a PRBS.
Zero-Valued Subcarrier
A subcarrier with no power. See also excluded subcarrier.
Zero-Valued Minislot
A minislot composed of zero-valued subcarriers and no pilots
20
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
4 ABBREVIATIONS AND ACRONYMS

This specification uses the following abbreviations:
s
Microsecond
ANSI
American National Standards Institute
AWGN
additive white Gaussian noise
BCH
Bose, Ray-Chaudhuri, Hocquenghem [codes]
BER
1) bit error ratio; 2) bit error rate
BPSK
binary phase shift keying
BW
bandwidth
CableLabs
Cable Television Laboratories, Inc.
CEA
Consumer Electronics Association
ceil
ceiling
CENELEC
European Committee for Electrotechnical Standardization
CER
codeword error ratio
CL
1) convergence layer; 2) CableLabs
CM
cable modem
CMCI
cable modem-to-customer premises equipment interface
CMTS
cable modem termination system
CNR
carrier to noise ratio
CP
1) cyclic prefix; 2) complementary pilot
CPE
customer premises equipment
CPU
central processing unit
CRC
cyclic redundancy check
CS
cyclic suffix
CSO
composite second order
CTB
composite triple beat
CW
1) continuous wave; 2) codeword
dB
decibel
dBc
decibel carrier
dBmV
decibel millivolt
dBr
decibel reference
DC
direct current
DCID
downstream channel identifier
DEPI
downstream external PHY interface
DFT
discrete Fourier transform
DHCP
dynamic host configuration protocol
DLS
DOCSIS light sleep [mode]
DOCSIS
DOCSIS 1.x
Data-Over-Cable Service Interface Specifications version 1.0 or 1.1
DOCSIS 2.0
Data-Over-Cable Service Interface Specifications version 2.0
10/29/13
CableLabs
21
CM-SP-PHYv3.1-I01-131029
DOCSIS 3.0
DOCSIS 3.1
DRFI
[DOCSIS] Downstream Radio Frequency Interface [Specification]
DRW
dynamic range window
DS
downstream
DSG
DOCSIS Set-top Gateway [Interface Specification]
DTI
DOCSIS Timing Interface [Specification]
DVB
Digital Video Broadcasting [Project]
DVB-C2
"Digital Video Broadcasting (DVB); Frame structure channel coding and modulation for a
second generation digital transmission system for cable systems (DVB-C2)"
eDOCSIS
embedded Data Over Cable Service Interface Specifications
EM
energy management [message]
EMC
electromagnetic compatibility
EN
European Standard (Europen Norme)
EQAM
edge quadrature amplitude modulation [modulator]
ERMI
Edge Resource Manager Interface [Specification]
ETSI
European Telecommunications Standards Institute
FCC
Federal Communications Commission
FDM
frequency division multiplexing
FDMA
frequency division multiple access
FEC
forward error correction
FFT
fast Fourier transform
FIR
finite impulse response
FR
fine ranging
ft
1) foot; 2) feet
GB
[Chinese] national standard (guobiao)
GB/T
[Chinese] recommended national standard (guobiao tuijian)
GF
Galois field
GHz
gigahertz
GT
guard time
HFC
hybrid fiber/coax
HRC
harmonic related carriers
Hz
hertz
in-phase
ICI
inter-carrier interference
I-CMTS
integrated cable modem termination system
ID
identifier
IDFT
inverse discrete Fourier transform
IEC
International Electrotechnical Commission
IETF
Internet Engineering Task Force
IFFT
inverse fast Fourier transform
IP
Internet protocol
22
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
IPDR
Internet protocol detail record
IPv4
Internet protocol version 4
IPv6
Internet protocol version 6
IR
initial ranging
IRC
incremental related carriers
ISI
inter-symbol interference
ISO
International Organization for Standardization
ITU
International Telecommunication Union
ITU-T
ITU Telecommunication Standardization Sector
kb
kilobit
kHz
kilohertz
L2VPN
layer 2 virtual private network
LAN
local area network
LDPC
low-density parity check
LFSR
linear feedback shift register
LLR
log-likelihood ratio
log
logarithm
LSB
least significant bit
LTE
long term evolution
MAC
media access control
MB
message block
MC
message channel
M-CMTS
modular cable modem termination system
MER
modulation error ratio
MHz
megahertz
ms
millisecond
MSB
most significant bit
MTC
multiple transmit channel [mode]
MULPI
MAC and upper layer protocols interface
NCP
next codeword pointer
NMS
network management system
ns
nanosecond
NSI
network side interface
NTSC
National Television System Committee
OFDM
orthogonal frequency division multiplexing
OFDMA
orthogonal frequency division multiple access
OOB
out-of-band
OSSI
operations support system interface
pilot
PAPR
peak-to-average power ratio
PDU
protocol data unit
10/29/13
CableLabs
23
CM-SP-PHYv3.1-I01-131029
PER
packet error ratio
PHY
physical layer
pk-pk
peak-to-peak
Pkt
packet
PLC
PHY link channel
PN
pseudorandom number
PRBS
pseudo-random binary sequence
Pre-eq
pre-equalization
ps
picosecond
PSD
power spectral density
Ptr
pointer
quadrature
QAM
quadrature amplitude modulation
QC-LDPC
quasi-cyclic low-density parity check
QoS
quality of service
QPSK
quadrature phase shift keying
RC
raised cosine
REQ
request
RF
radio frequency
RFC
request for comments
RFI
radio frequency interface
RFoG
radio frequency over glass
RMS
root mean square
RP
roll-off prefix
R-S
Reed-Solomon
RX
1) receive; 2) receiver
second
SAC
Standardization Administration of the People's Republic of China
S-CDMA
synchronous code division multiple access
SCN
signal-to-composite noise [ratio]
SC-QAM
single carrier quadrature amplitude modulation
SCTE
Society of Cable Telecommunications Engineers
SEC
security
SID
service identifier
SNMP
simple network management protocol
SNR
signal-to-noise ratio
TCM
Trellis coded modulation
TCS
transmit channel set
TDM
time division multiplexing
TDMA
time division multiple access
TEI
TDM emulation interface
24
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
TS
time stamp
TV
television
TX
1) transmit; 2) transmitter
UCD
upstream channel descriptor
UGHB
under-grant hold bandwidth
UGHU
under-grant hold number of users
UID
unique identifier
URL
uniform resource locator
US
upstream
XOR
exclusive or
10/29/13
CableLabs
25
CM-SP-PHYv3.1-I01-131029
5 OVERVIEW AND FUNCTIONAL ASSUMPTIONS

This section describes the characteristics of a cable television plant, assumed to be for the purpose of operating a
data-over-cable system.
The cable plants have very diverse physical topologies. These topologies range from fiber to the home node
architectures as well as fiber nodes with many actives in cascade. The plant characteristics described in this section
covers the great majority of plant scenarios.
This section is not a description of CMTS or CM parameters. The data-over-cable system MUST be interoperable
within the environment described in this section.
Whenever a reference in this section to frequency plans, or compatibility with other services, conflicts with any legal
requirement for the area of operation, the latter shall take precedence. Any reference to NTSC analog signals in 6
MHz channels does not imply that such signals are physically present.
5.1
Overview
This specification defines the PHY layer protocol of DOCSIS 3.1. It also describes the channel assumptions over
which DOCSIS 3.1 systems are expected to operate.
DOCSIS 3.1 ultimate service goal of multi-gigabit per second in the downstream direction and gigabit per second in
the upstream direction resulted in significant changes in the PHY layer approach compared to earlier DOCSIS
versions in addition to changes on the cable network assumptions. DOCSIS 3.1 focuses on the eventual use of the
entire spectrum resources available in cable environment by the CMTS and CM and on scalable cost effective
techniques to achieve full spectrum use.
DOCSIS 3.1 assumes OFDM downstream signals and OFDMA upstream signals to achieve robust operation and
provide more efficient use of the spectrum than previous DOCSIS versions. In order to reach the target service goal
in the upstream direction, plant changes on the upstream/downstream spectrum split are expected. The DOCSIS 3.1
system will have options of several split configurations that can be exercised based on traffic demand, services
offered and the capability of the cable plant. In the downstream direction, HFC plant spectrum extended beyond the
current 1002 MHz is expected.
The DOCSIS 3.0 systems and earlier versions are sometimes referred to in this document as single carrier QAM
(SC-QAM) systems in contrast to the multicarrier DOCSIS 3.1 OFDM/OFDMA system.
The OFDM downstream multicarrier system is composed of a large number of subcarriers that have either 25 kHz or
50 kHz spacing. These subcarriers are grouped into independently configurable OFDM channels each occupying a
spectrum of up to 192 MHz in the downstream, totaling 7680 25 kHz subcarriers or 3840 50 kHz subcarriers; of
which up to 7601 (25 kHz) or 3801 (50 kHz) active subcarriers span 190 MHz. The OFDMA upstream multicarrier
system is also composed of either 25 kHz or 50 kHz subcarriers. In the upstream, the subcarriers are grouped into
independently configurable OFDMA channels each occupying a spectrum of up to 96 MHz, totaling 3840 25 kHz
spaced subcarriers or 1920 50 kHz spaced subcarriers. Many parameters of these channels can be independently
configured thereby optimizing configuration based on channel conditions.
The cable topologies have been evolving to smaller node architectures with fewer amplifiers in cascade. This natural
cable network evolution and gradual reduction in node sizes bring a corresponding improvement in channel
conditions. A DOCSIS 3.1 goal is to leverage the expected network improvements and achieve higher efficiency
levels corresponding to improvement in channel conditions. In the downstream and in the upstream directions,
profiles can be defined to match the transmission configuration to the channel conditions with greater granularity.
DOCSIS 3.1 technology is able to operate in classic cable network topologies, but those networks may not be able to
achieve full capabilities of DOCSIS 3.1 bandwidth efficiencies.
An assumption in the topology configuration of DOCSIS 3.1 is that the CM is predominantly placed in a gateway
architecture configuration. Specifically the CM is located either at the drop-home boundary or after the first twoway splitter within the customer premises. This configuration limits potential attenuation within the home
environment. This reduced attenuation is leveraged to enable higher efficiencies intended in DOCSIS 3.1.
26
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Another intent of DOCSIS 3.1 is to leverage the granularity from multiple narrowband subcarriers to exclude
unwanted spectrum regions from transmission so that interferers can be avoided. Leveraging the same mechanism,
coexistence with existing systems can be implemented by carving out a portion of the spectrum to allow for another
transmission to co-exist.
It is expected that CMs and CMTS are able to operate in multiple modes. A mode could be pure SC-QAM, pure
OFDM transmission or a combination of the two. This flexibility enables a smooth transition between SC-QAM and
OFDM systems.
DOCSIS 3.1 uses Low-Density-parity-Check (LDPC) FEC coding instead of the Reed-Solomon used in DOCSIS
3.0 and earlier versions. Long FEC codeword types are defined in the upstream and downstream to optimize
efficiency. LDPC FEC along with frequency and time interleaving is used to provide robustness against narrowband
interferers and burst events.
In several instances equivalent characterization and metrics to those used in DOCSIS 3.0 and earlier versions have
been devised to facilitate comparison of requirements among different versions of the specification.
The DOCSIS 3.1 suite of specifications includes considerations to improve and optimize operation under special
modes. One mode is a light sleep mode that is introduced to minimize CM power consumption. Also, some features
are defined primarily operation in FTTH network architectures configured for RFoG.
5.2
Functional Assumptions
5.2.1
Equipment Assumptions
5.2.1.1
Frequency Plan
In the downstream direction, the cable system is assumed to have a pass band with a lower edge of either 54 MHz,
87.5 MHz, 108 MHz or 258 MHz, and an upper edge that is implementation-dependent but is typically in the range
of 550 to 1002 MHz. Upper frequency edges extending to 1218 MHz, 1794 MHz and others are expected in future
migrations of the plants. Within that pass band, NTSC analog television signals in 6 MHz channels are assumed
present on the standard, HRC or IRC frequency plans of [CEA-542-D], as well as other narrowband and wideband
digital signals.
In the upstream direction, the cable system may have a 5-42 MHz, 5-65 MHz, 5-85 MHz, 5-117, 5-204 MHz or pass
bands with an upper band edge beyond 204 MHz. NTSC analog television signals in 6 MHz channels may be
present, as well as other signals.
5.2.1.2
Compatibility with Other Services
The CM MUST coexist with any services on the cable network.

The CMTS MUST coexist with any services on the cable network.
In particular:
The CMTS MUST be interoperable in the cable spectrum assigned for CMTS and CM interoperation while the
balance of the cable spectrum is occupied by any combination of television and other signals.
The CM MUST be interoperable in the cable spectrum assigned for CMTS and CM interoperation while the
balance of the cable spectrum is occupied by any combination of television and other signals.
The CMTS MUST NOT cause harmful interference to any other services that are assigned to the cable network
in spectrum outside of that allocated to the CMTS.
The CM MUST NOT cause harmful interference to any other services that are assigned to the cable network in
spectrum outside of that allocated to the CM.
Harmful interference is understood as:
No measurable degradation (highest level of compatibility),
10/29/13
CableLabs
27
CM-SP-PHYv3.1-I01-131029
No degradation below the perceptible level of impairments for all services (standard or medium level of
compatibility), or
No degradation below the minimal standards accepted by the industry (for example, FCC for analog video
services) or other service provider (minimal level of compatibility).
5.2.1.3
Fault Isolation Impact on Other Users
As CMTS transmissions are on a shared-media, point-to-multipoint system, fault-isolation procedures should take
into account the potential harmful impact of faults and fault-isolation procedures on numerous users of the dataover-cable, video and other services.
For the interpretation of harmful impact, see Section 5.2.1.2 above.
5.2.1.4
Cable System Terminal Devices
The CM is expected to meet and preferably exceed all applicable regulations for Cable System Termination Devices
and Cable Ready Consumer Equipment as defined in FCC Part 15 [FCC15] and Part 76 [FCC76]. None of these
specific requirements may be used to relax any of the specifications contained elsewhere within this document.
5.2.2
RF Channel Assumptions
The data-over-cable system, configured with at least one set of defined physical-layer parameters (e.g., modulation,
interleaver depth, etc.) from the range of configurable settings described in this specification, is expected to be
interoperable on cable networks having characteristics defined in this section. This is accomplished in such a manner
that the forward error correction provides for equivalent operation in a cable system both with and without the
impaired channel characteristics described below.
5.2.2.1
Transmission Downstream
The RF channel transmission characteristics of the cable network in the downstream direction are described in Table
51. These numbers assume total average power of a digital signal in a 192 MHz channel bandwidth for subcarrier
levels unless indicated otherwise. For impairment levels, the numbers in Table 51 assume average power in a
bandwidth in which the impairment levels are measured in a standard manner for cable TV systems. For analog
signal levels, the numbers in Table 51 assume peak envelope power in a 6 MHz channel bandwidth. All conditions
are present concurrently. It is expected that the HFC plant will have better conditions for DOCSIS 3.1 to provide the
higher throughput and capacities anticipated.
Table 51 - Typical Downstream RF Channel Transmission Characteristics
Parameter
Value
Frequency range
Cable system normal downstream operating range is from

54 MHz to 1002 MHz. Extended operating ranges include
lower downstream edges of 108 MHz and 258 MHz and
upper downstream edges of 1218 MHz and 1794 MHz.
RF channel spacing (design bandwidth)
24 to 192 MHz
One way transit delay from headend to most distant

customer
0.800 ms (typically much less)
Signal to Composite Noise Ratio
35 dB
Carrier-to-Composite triple beat distortion ratio
Not less than 41 dB
Carrier-to-Composite second order distortion ratio
Not less than 41 dB
Carrier-to-Cross-modulation ratio
Not less than 41 dB
Carrier-to-any other discrete interference (ingress)
Not less than 41 dB
Maximum amplitude variation across the 6 MHz

channel (digital channels)
1.74 dB pk-pk/6 MHz
28
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Parameter
Value
113 ns over 24 MHz
-20 dBc @ 0.5 s
-25 dBc @ 1.0 s
-30 dBc @ 1.5 s
-35 dBc @ > 2.0 s
-40 dBc @ > 3.0 s
-45 dBc @ > 4.5 s
-50 dBc @ > 5.0 s

Micro-reflections bound for dominant single echo
Carrier hum modulation
Not greater than -30 dBc (3%)
Maximum analog video carrier level at the CM input
17 dBmV
Maximum number of analog carriers
121
5.2.2.2
Transmission Upstream
The RF channel transmission characteristics of the cable network in the upstream direction are described in Table 5
2. No combination of the following parameters will exceed any stated interface limit defined elsewhere in this
specification. Transmission is from the CM output at the customer location to the headend.
Table 52 - Typical Upstream RF Channel Transmission Characteristics
Parameter
Value
Frequency range
Cable system normal upstream operating range is from

a lower band-edge of 5 MHz to upper band-edges to 42
MHz and 65 MHz. Extended operating ranges include
upper upstream band-edges of 85 MHz, 117 MHz and
204 MHz.
One way transit delay from most distant customer to

headend.
0.800 ms (typically much less)
Carrier-to-interference plus ingress (the sum of noise,

distortion, common-path distortion and cross modulation
and the sum of discrete and broadband ingress signals,
impulse noise excluded) ratio
Not less than 25 dB
Carrier hum modulation
Not greater than -26 dBc (5.0%)
Maximum amplitude variation across the 6 MHz channel

(digital channels)
2.78 dB pk-pk/6 MHz
163 ns over 24 MHz

-16 dBc @ 0.5 s
-22 dBc @ 1.0 s
-29 dBc @ 1.5 s
-35 dBc @ > 2.0 s
-42 dBc @ > 3.0 s
-51 dBc @ > 4.5 s
Micro-reflections bound for dominant single echo
Seasonal and diurnal reverse gain (loss) variation
5.2.2.2.1
Not greater than 14 dB min to max
Availability
Cable network availability is typically greater than 99.9%.
10/29/13
CableLabs
29
CM-SP-PHYv3.1-I01-131029
5.2.3
Transmission Levels
The nominal power level of the upstream CM signal(s) will be as low as possible to achieve the required margin
above noise and interference. Uniform power loading per unit bandwidth is commonly followed in setting upstream
signal levels, with specific levels established by the cable network operator to achieve the required carrier-to-noise
and carrier-to-interference ratios.
5.2.4
Frequency Inversion
There will be no frequency inversion in the transmission path in either the downstream or the upstream directions,
i.e., a positive change in frequency at the input to the cable network will result in a positive change in frequency at
the output.
30
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
6 PHY SUBLAYER FOR SC-QAM

6.1
Scope
This section applies to cases where a DOCSIS 3.1 CM or CMTS is operating with SC-QAM operation only, with no
OFDM operation. As such, it represents backward compatibility requirements when operating with DOCSIS 3.0
systems or with the new DOCSIS 3.1 PHY disabled. It also applies only to the first technology option referred to in
Section 1.1; for the second option refer to Annex C; and for the third option refer to Annex D.
This specification defines the electrical characteristics and signal processing operations for a CM and CMTS. It is
the intent of this specification to define an interoperable CM and CMTS such that any implementation of a CM can
work with any CMTS. It is not the intent of this specification to imply any specific implementation.
As the requirements for a DOCSIS 3.1 CM and CMTS are largely unchanged relative to DOCSIS 3.0 devices for
SC-QAM operation, this section is comprised primarily of references to the appropriate DOCSIS 3.0 specification
sections for the specific requirements for a DOCSIS 3.1 CM and CMTS, as well as any deltas relative to those
requirements (the primary difference being that a DOCSIS 3.1 CM and CMTS are required to support a minimum of
24 downstream and 8 upstream channels instead of 4 downstream and 4 upstream as in DOCSIS 3.0 devices).
A DOCSIS 3.1 CM MUST comply with the referenced requirements in the PHYv3.0 and DRFI specifications noted
in this section, with the exception of any deltas called out in this section (which will be identified with separate
requirement statements). A DOCSIS 3.1 CMTS MUST comply with the referenced requirements in the PHYv3.0
and DRFI specifications noted in this section, with the exception of any deltas called out in this section (which will
be identified with separate requirement statements).
6.2
Upstream Transmit and Receive
This section is based on section 6.2 of [DOCSIS PHYv3.0].

6.2.1
Overview
See section 6.2.1 of [DOCSIS PHYv3.0], with the exception noted below.
A CM MUST support at least eight (8) active upstream channels (which are referred to as the Transmit Channel Set
for that CM).
A CMTS MUST support at least eight (8) active upstream channels.
A CMTS SHOULD support S-CDMA mode.
6.2.2
Signal Processing Requirements
See section 6.2.2 of [DOCSIS PHYv3.0].

6.2.3
Modulation Formats

6.2.4
R-S Encode

6.2.5
Upstream R-S Frame Structure (Multiple Transmit Enabled)

6.2.6
Upstream R-S Frame Structure (Multiple Transmit Disabled)
10/29/13
CableLabs
31
CM-SP-PHYv3.1-I01-131029
6.2.7
TDMA Byte Interleaver

6.2.8
Scrambler (randomizer)

6.2.9
TCM Encoder

6.2.10 Preamble Prepend
6.2.11 Modulation Rates
6.2.12 S-CDMA Framer and Interleaver
6.2.13 S-CDMA Framer
6.2.14 Symbol Mapping
6.2.15 S-CDMA Spreader
6.2.16 Transmit Pre-Equalizer
6.2.17 Spectral Shaping
6.2.18 Relative Processing Delays
6.2.19 Transmit Power Requirements
6.2.20 Burst Profiles
6.2.21 Burst Timing Convention
6.2.22 Fidelity Requirements
32
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
6.2.23 Upstream Demodulator Input Power Characteristics

6.2.24 Upstream Electrical Output from the CM
6.2.25 Upstream CM Transmitter Capabilities
6.3
Downstream Transmit
This section is based on section 6.3 of [DOCSIS DRFI].

6.3.1
Downstream Protocol
See section 6.3.1 of [DOCSIS DRFI].

6.3.2
Spectrum Format

6.3.3
Scaleable Interleaving to Support Video and High-Speed Data Services

6.3.4
Downstream Frequency Plan

6.3.5
DRFI Output Electrical
For legacy, SC-QAMs, the EQAM and CMTS MUST support the electrical output requirements specified in the
following sections and tables of [DOCSIS DRFI]:
Section 6.3.5, DRFI Output Electrical
Section 6.3.5.1, CMTS or EQAM Output Electrical
Table 6-3, RF Output Electrical Requirements
Section 6.3.5.1.1, Power per Channel CMTS or EQAM
Table 6-4, DRFI Device Output Power
Section 6.3.5.1.2, Independence of Individual Channels within the Multiple Channels on a Single RF Port
Section 6.3.5.1.3, Out-of-Band Noise and Spurious Requirements for CMTS or EQAM
Table 6-5, EQAM or CMTS Output Out-of-Band Noise and Spurious Emissions Requirements for N =< 8
Table 6-6, EQAM or CMTS Output Out-of-Band Noise and Spurious Emissions Requirements N>=9 and
N'>=N/4
Table 6-7, EQAM or CMTS Output Out-of-Band Noise and Spurious Emissions Requirements N>=9 and
N'<N/4
6.3.6
CMTS or EQAM Clock Generation
10/29/13
CableLabs
33
CM-SP-PHYv3.1-I01-131029
6.3.7
Downstream Symbol Clock Jitter for Synchronous Operation

6.3.8
Downstream Symbol Clock Drift for Synchronous Operation

6.3.9
Timestamp Jitter
6.4
Downstream Receive
This section is based on section 6.3 of [DOCSIS PHYv3.0].

6.4.1
Downstream Protocol and Interleaving Support

6.4.2
Downstream Electrical Input to the CM
See section 6.2.1 of [DOCSIS PHYv3.0], with the exception noted below.
A CM MUST support at least 24 active downstream channels.
A CMTS MUST support at least 24 active downstream channels.
6.4.3
CM BER Performance

6.4.4
Downstream Multiple Receiver Capabilities

6.4.5
Non-Synchronous DS Channel Support
34
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
7 PHY SUBLAYER FOR OFDM

7.1
Scope
This specification defines the electrical characteristics and signal processing operations for a cable modem (CM) and
Cable Modem Termination System (CMTS). It is the intent of this specification to define an interoperable CM and
CMTS such that any implementation of a CM can work with any CMTS. It is not the intent of this specification to
imply any specific implementation.
This section describes cable modem and CMTS physical layer requirements for orthogonal frequency division
multiplexing (OFDM) modulation of RF signals transmitted between the modem and CMTS. This mode of
operation is separate and distinct from single-carrier QAM operation described in Section 6.
Section 7 is organized into subsections as summarized below:
7.1 Scope
Describes the structure of the DOCSIS 3.1 Physical Layer (PHY) specification chapter on OFDM operation.
7.2 Upstream and Downstream Frequency Plan
Defines requirements for the range of frequencies in the RF spectrum used for transmission of information from
the cable modem to the CMTS (upstream) and for the range of frequencies in the RF spectrum used for the
transmission of information from the CMTS to the cable modem (downstream).
7.3 OFDM Numerology
Lists the key OFDM parameters for each of the two configurable options for FFT/IFFT point values.
Information includes the specified master clock frequency, sampling rate, channel bandwidths, subcarrier
spacing and other parameters defining OFDM operation.
7.4 Upstream Transmit and Receive
Defines requirements for electrical characteristics and signal processing operations for the CM transmitting
OFDM-modulated RF subcarriers upstream to the CMTS and for the CMTS receiving OFDM-modulated RF
subcarriers from CMs.
7.5 Downstream Transmit and Receive
Defines requirements for electrical characteristics and signal processing operations for the CMTS transmitting
OFDM-modulated RF subcarriers downstream to CMs and for CMs receiving OFDM-modulated RF subcarriers
transmitted by the CMTS.
7.2
Upstream and Downstream Frequency Plan
The following spectrum definitions are based on the system requirement that the downstream transmission
frequencies always reside above the upstream transmission frequencies in the cable plant.
7.2.1
Downstream CM Spectrum
The CM MUST support a minimum of two independently configurable OFDM channels each occupying a spectrum
of up to 192 MHz in the downstream.
The CM MUST support a downstream upper band edge of 1.218 GHz.
The CM MAY support a downstream upper band edge of 1.794 GHz.
The CM SHOULD support a downstream lower band edge of 108 MHz when the CM is configured to use an
upstream upper band edge of 85 MHz or less.
7.2.2
Downstream CMTS Spectrum
The CMTS MUST support a minimum of two independently configurable OFDM channels each occupying a
spectrum of up to 192 MHz in the downstream.
The CMTS MUST support a downstream upper band edge of 1.218 GHz.
10/29/13
CableLabs
35
CM-SP-PHYv3.1-I01-131029
The CMTS MAY support a downstream upper band edge of 1.794 GHz.
The CMTS MUST support a downstream lower band edge of 258 MHz.
The CMTS SHOULD support a downstream lower band edge of 108 MHz.
7.2.3
Upstream CM Spectrum
The CM MUST support a minimum of two independently configurable OFDMA channels each occupying a
spectrum of up to 96 MHz in the upstream.
The CM MAY support more than two independently configurable OFDMA channels each occupying a spectrum of
up to 96 MHz in the upstream.
The CM MUST support upstream transmissions from 5 to at least 204 MHz and agile placement of the OFDMA
channels within that range.
Individual CM implementations may limit the spectrum over which the CM is able to transmit upstream signals. As
a result, in order to be compliant with this specification a CM MUST support one or more of the following upstream
upper band edges, as long as one of the upstream upper band edges supported is 85 MHz or greater: 42 MHz; 65
MHz, 85 MHz, 117 MHz, and/or 204 MHz.
The CM MUST be configurable to operate with any supported upstream upper band edge.
The CM MAY support additional spectrum beyond 204 MHz for the upstream.
The CM MUST NOT cause harmful interference to any downstream signals that might exist above its configured
upstream upper band edge.
The CM MUST be capable of transmitting 192 MHz of active channels when operating with the 204 MHz upstream
upper band edge.
In DOCSIS 3.1 upstream mode the CM MUST be capable of transmitting OFDMA channels and legacy SC-QAM
channels at the same time (as controlled by the CMTS). In all cases the CM is not required to transmit legacy SCQAM channels above a frequency of 85 MHz.
7.2.4
Upstream CMTS Spectrum
The CMTS MUST support a minimum of two independently configurable OFDMA channels each occupying a
spectrum of up to 96 MHz in the upstream.
The CMTS MAY support more than two independently configurable OFDMA channels each occupying a spectrum
of up to 96 MHz in the upstream.
The CMTS MUST support upstream transmissions from 5 to at least 204 MHz and agile placement of the OFDMA
blocks within that range.
The CMTS MAY support additional spectrum beyond 204 MHz for the upstream.
The CMTS MUST capable of receiving 192 MHz of active channels when operating with the 204 MHz upstream
upper band edge. In DOCSIS 3.1 upstream mode the CM is capable of transmitting OFDMA channels and legacy
SC-QAM channels at the same time (as controlled by the CMTS). In all cases the CMTS MUST NOT configure the
CM to transmit legacy SC-QAM channels above a frequency of 85 MHz.
7.2.5
Exclusion Band Rules
During OFDM/OFDMA channel planning, the following rules are to be observed to ensure proper operation of
DOCSIS 3.1 CMTS and CM.
The CMTS MUST ensure that the configured OFDM/OFDMA channels are aligned with the rules specified in
Sections 7.2.5.1 and 7.2.5.2.
36
CableLabs
10/29/13
7.2.5.1
CM-SP-PHYv3.1-I01-131029
Downstream Exclusion Band Rules
There has to be at least one contiguous modulated OFDM bandwidth of 22 MHz or greater, which will enable
an OFDM channel bandwidth of 24 MHz including guardbands.
Exclusion bands separate contiguous modulation bands.
The minimum contiguous modulation band has to be 2 MHz.
Exclusion bands and individually excluded subcarriers are common to all downstream profiles.
Exclusion bands are a minimum of 1 MHz but increment above 1 MHz by granularity of individual subcarrier
(25 kHz for 8k FFT and 50 kHz for 4K FFT).
The ONLY exception to the above is for exclusion bands that are allowed to occupy the following frequency
ranges in alignment with FCC regulations.
121.400 MHz to 121.600 MHz
156.750 MHz to 156.850 MHz
242.950 MHz to 243.050 MHz
405.925 MHz to 406.176 MHz
Unique spurious emissions requirements exist for these bands separate from the general exclusion bands
requirements.
Exclusion bands plus individually excluded subcarriers are limited to 20% or less of spanned modulation
spectrum, where the spanned modulation spectrum is defined as: frequency of maximum active subcarrier
frequency of minimum active subcarrier.
The number of individually excluded subcarriers is limited by the following:
The total spectrum of individually excluded subcarriers cannot exceed 5% of any contiguous modulation
spectrum.
The total spectrum of individually excluded subcarriers cannot exceed 5% of a 6 MHz moving window
across the contiguous modulation spectrum.
The total spectrum of individually excluded subcarriers cannot exceed 20% of a 1 MHz moving window
across the contiguous modulation spectrum.
The 6 MHz of contiguous spectrum reserved for the PLC cannot have any exclusion bands or excluded
subcarrier.
7.2.5.2
Upstream Exclusion Band Rules
Excluded subcarriers will only occur between minislots.
Subcarrier exclusions between minislots can be any integer number of subcarriers. There is no minimum
subcarrier exclusion restriction.
7.3
OFDM Numerology
7.3.1
Downstream OFDM Numerology
DOCSIS 3.1 uses OFDM for downstream modulation.

Two modes of operation are defined for the downstream: 4K FFT and 8K FFT modes for a sampling rate of
204.8 MHz. These are described in Section 7.5.7.
Table 71 summarizes the numerical values for the downstream OFDM parameters; a more detailed description of
the parameters is given in the sections which follow.
10/29/13
CableLabs
37
CM-SP-PHYv3.1-I01-131029
Table 71 - Downstream OFDM parameters

Parameter
4K mode
8K mode
Downstream master clock frequency
10.24 MHz
Downstream Sampling Rate (fs)
204.8 MHz
Downstream Elementary Period (Tsd)
1/(204.8 MHz)
Channel bandwidths
24 MHz 192 MHz

2 MHz
Minimum contiguous modulation band (see Section

7.2.5.1)
IDFT size
4096
8192
50 kHz
25 kHz
FFT duration (Useful symbol duration) (Tu)
20 s
40 s
Number of active subcarriers in signal (192 MHz channel)

Values refer to 190 MHz of used subcarriers.
3801
7601
Subcarrier spacing
Spacing between first and last active subcarrier
190 MHz
Cyclic Prefix
0.9375 s
1.25 s
2.5 s
3.75 s
5 s
(192 * Tsd)
(256 * Tsd)
(512 * Tsd)
(768 * Tsd)
(1024 * Tsd)
Windowing
Tukey raised cosine window, embedded into cyclic prefix

0 s
(0 * Tsd)
0.3125 s
(64 * Tsd)
0.625 s
(128 * Tsd)
0.9375 s
(192 * Tsd)
1.25 s
( 256 * Tsd)
The downstream OFDM channel bandwidth can be any value from 24 MHz to 192 MHz. Smaller bandwidths than
192 MHz are achieved by nulling subcarriers prior to the IDFT, i.e., by adjusting the equivalent number of active
subcarriers while maintaining the same subcarrier spacing of 25 kHz or 50 kHz.
7.3.2
Upstream OFDMA Numerology
DOCSIS 3.1 uses OFDMA (orthogonal frequency-division multiple access) for upstream modulation. OFDMA is a
multi-user version of OFDM, and assigns subsets of subcarriers to individual CMs. The upstream OFDMA
parameters are derived from the downstream parameters, and are summarized in Table 72. A more detailed
description of the parameters is given in the sections which follow.
Table 72 - Upstream OFDMA Parameters
Parameter
2k Mode
102.4 MHz
Upstream Sampling Rate (fsu)
1/102.4 MHz
Upstream Elementary Period Rate (Tsu)
10 MHz, , 96 MHz
6.4 MHz, , 96 MHz
2048
4096
50 kHz
25 kHz
FFT duration (Useful symbol duration) (Tu)
20 s
40 s
Number of active subcarriers in signal (96MHz channel)

Values refer to 95 MHz of active subcarriers
1901
3801
Channel bandwidths
IDFT size
(depending on channel bandwidth)
Subcarrier spacing
38
4k Mode
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Parameter
2k Mode
4k Mode
Upstream Cyclic Prefix
0.9375 s
1.25 s
1.5625 s
1.875 s
2.1875 s
2.5 s
2.8125 s
3.125 s
3.75 s
5.0 s
6.25 s
Upstream window size
Tukey raised cosine window, embedded into cyclic prefix

0 s
(0 * Tsu)
0.3125 s
(32 * Tsu)
0.625 s
(64 * Tsu)
0.9375 s
(96 * Tsu)
1.25 s
(128 * Tsu)
1.5625 s
(160 * Tsu)
1.875 s
(192 * Tsu)
2.1875 s
(224 * Tsu)
7.3.3
(96 * Tsu)
(128 * Tsu)
(160 * Tsu)
(192 * Tsu)
(224 * Tsu)
(256 * Tsu)
(288 * Tsu)
(320 * Tsu)
(384 * Tsu)
(512 *Tsu)
(640 *Tsu)
Subcarrier Clocking
The "locking" of subcarrier "clock and carrier" are defined and characterized by the following rules:
Each OFDM symbol is defined with a Subcarrier Clock frequency of nominally 20 usec or 40 usec. For each
OFDM symbol, the Subcarrier Clock period (us) may vary from nominal with limits defined in Section 7.5.3.
The number of cycles of each subcarrier generated by the CMTS during one period of the Subcarrier Clock (for
each OFDM symbol) MUST be an integer number.
The CMTS Subcarrier Clock MUST be synchronous with the 10.24 MHz Master Clock defined by:
Subcarrier Clock frequency = (M/N)*Master Clock frequency where M = 20 or 40, and N = 8192
The limitation on the variation from nominal of the Subcarrier Clock frequency at the output connector is
defined in Section 7.5.3.
Each OFDM symbol has a cyclic prefix which is an integer multiple of 1/64th, of the Subcarrier Clock period.
Each OFDM symbol duration is the sum of one Subcarrier Clock period and the cyclic prefix duration.
The number of cycles of each subcarrier generated by the CMTS during the OFDM symbol duration (of each
symbol) MUST be K+K*L/64, where K is an integer related to the subcarrier index and frequency upconversion
of the OFDM channel, and L is an integer related to the cyclic prefix. (K is an integer related to the subcarrier
index and increases by 1 for each subcarrier).
The phase of each subcarrier within one OFDM symbol is the same, when each is assigned the same
constellation point (I + jQ), relative to the Reference Time of the OFDM symbol. There is nominally no change
in phase on each subcarrier for every cycle of 64 OFDM symbols, when both are assigned the same I + jQ, and
referenced to the Reference Time of their respective OFDM symbol.
7.4
Upstream Transmit and Receive
7.4.1
Overview
This section specifies the upstream transmission electrical and signal processing requirements for the transmission
of OFDM modulated RF signals from the CM to the CMTS.
10/29/13
CableLabs
39
CM-SP-PHYv3.1-I01-131029
This section is organized into subsections as summarized below:

7.4.1 Overview
Describes the structure of the DOCSIS 3.1 Physical Layer (PHY) specification sub-section on CM transmitter
and CMTS receiver requirements.
7.4.2 Signal Processing and Requirements
Informative description of DOCSIS 3.1 upstream channel signal processing including descriptions of data
framing and subcarrier structure, forward error correction, bit randomization, mapping of bits to constellation
symbols, interleaving, pre-equalization, inverse Fourier transformation, and pre-pending the cyclic prefix.
7.4.3 Time and Frequency Synchronization
Defines the requirements for synchronizing the transmission of subcarriers to the system clock.
7.4.4 Forward Error Correction
Defines the requirements for upstream forward error correction codewords and encoding.
7.4.5 Data Randomization
Defines the requirements for randomization of codewords.
7.4.6 Time and Frequency Interleaving and De-interleaving
Defines the requirements for time and frequency interleaving and de-interleaving of subcarriers for a grant.
7.4.7 Mapping of Bits to Cell Words
Defines, CM requirements for mapping serialized data bits from the randomizer to QAM constellations and
CMTS requirements for recovering bits from cell words.
7.4.8 Mapping and De-mapping Bits to/from QAM subcarriers
Defines modulation formats, CM requirements for mapping data bits to subcarriers in minislots according to the
profile definition and CMTS requirements for recovering bits from QAM constellations.
7.4.9 REQ Messages
Defines CM requirements for creating REQ messages for requesting transmission opportunities from the
CMTS.
7.4.10 IDFT
Defines the Inverse Discrete Fourier Transform formula used by the CM to convert data from frequency domain
to time domain for upstream transmission.
7.4.11 Cyclic Prefix and Windowing
Defines CM requirements for applying cyclic prefix for mitigating interference due to micro-reflections and for
applying windowing to maximize the capacity of the upstream channel. Defines CMTS requirements for
supporting cyclic prefixes in upstream data.
7.4.12 OFDMA Fidelity Requirements
Defines the CM transmitter requirements and limits for transmit power, spurious emissions, phase noise and
modulation error ratio.
7.4.13 Cable Modem Transmitter Output Requirements
Defines CM transmitter output requirements including OFDMA parameters, output power level, impedance,
channel bandwidth and return loss.
7.4.14 CMTS Receiver Input Power Requirements
Defines CMTS receiver input power requirements.
7.4.15 Ranging
Defines the requirements for initial ranging, fine ranging, and wideband probing.
7.4.16 Upstream Pilot Structure
Defines CM requirements for transmitting pilots and CMTS requirements for receiving pilots enabling the CM
and CMTS to adapt to channel conditions and frequency offset.
40
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
7.4.17 Upstream Pre-Equalization

Defines CMTS and CM requirements for estimating the channel using probes and for modifying preequalization coefficients as needed during ranging.
7.4.2
Signal Processing and Requirements
Upstream transmission uses OFDMA frames. Each OFDMA frame is comprised of a configurable number of
OFDM symbols, K. Several transmitters may share the same OFDMA frame by transmitting data and pilots on
allocated sub-carriers of the OFDMA frame. There are several pilot patterns as described in Section 7.4.16.
The structure of an OFDMA frame is depicted in Figure 71. The upstream spectrum is divided into groups of subcarriers called minislots. Minislots have dedicated sub-carriers, all with the same modulation order ("bit loading"). A
CM is allocated to transmit one or more minislots in a Transmission Burst. The modulation order of a minislot, as
well as the pilot pattern to use may change between different transmission bursts and are determined by a
transmission profile.
minislot N-1
single minislot
TX Burst
minislot m+k-1
minislot m+k-1
k minislot
TX Burst
Minislot m
minislot m-1
m- minislot
TX Burst
minislot 1
Q
subcarriers
minislot 0
K symbols
Figure 71 - OFDMA Frame Structure
Serial data signals received from the PHY-MAC Convergence Layer are received and processed by the PHY as
illustrated in Figure 72. This process yields a transmission burst of a single or multiple OFDMA minislots, as
allocated by the PHY-MAC Convergence Layer. Each minislot is comprised of pilots, complementary pilots, and
data subcarriers, as described in Section 8.2.2. Subcarriers that are not used for data or pilots are set to zero.
Probe
Generator
Data from
PHY MAC
Interface
FEC
Padding
LDPC
Encoder
Scrambler
Symbol
Mapper
Interleaver
/ OFDMA
Framer
Bit
Loading
Pilot
Pattern
Pre-EQ
IFFT
CP and
Window
Figure 72 - Upstream transmitter block diagram
10/29/13
CableLabs
41
CM-SP-PHYv3.1-I01-131029
This section briefly describes the process and provides links to the specific requirements for each process described
in this specification.
7.4.2.1
Framing
Figure 73 describes how the received bits from the PHY-MAC Convergence layers are framed before being
converted into constellation symbols. The number of FEC padding bits and the number of minislot padding bits are
calculated by the PHY-MAC Convergence layer according to the allocation of minislot and the profile received by
the grant message.
Figure 73 - Upstream Transmitter Block Diagram
7.4.2.2
Forward Error Correction Encoding
Data received from the PHY-MAC Convergence layer interface, along with the FEC padding, is LDPC encoded.
The upstream has three LDPC codes: a long, medium, and short FEC code, as described in Section 7.4.4. Prior to
encoding, the transmitter must decide on the configuration of the codewords and shortening as described in Section
7.4.4.3. If required, FEC padding bits are added prior to the encoding.
7.4.2.3
Randomizer and Symbol Mapper
The encoded bits are then randomized (scrambled) using the PRBS scrambler as specified in Section 7.4.5. The
scrambler output bits are converted into constellation symbols according to the corresponding modulation order of
the minislot. All subcarriers of a given type (Pilots, Complementary Pilots, Data subcarriers, Null subcarriers) in a
minislot have the same modulation order. The Symbol Mapper is described in Section 7.4.7.
7.4.2.4
OFDMA framer and Interleaver
Constellation symbols then enter the OFDMA framer and Interleaver block. The OFDMA framer adds pilots
according to the pilot pattern associated with the transmission burst minislot. The constellation symbols are written
to subcarriers associated with the transmission burst minislots and are then interleaved in time and frequency as
described in Section 7.4.6. If required, idle bits are added to unused data subcarriers in the minislots.
7.4.2.5
Pre-Equalization
The upstream transmitter applies pre-equalization as described in Section 7.4.17 in order to pre-distort the
transmitted signals according to coefficients received from the CMTS to compensate for the channel response.
7.4.2.6
IFFT Transformation
In this stage each pre-equalized symbol is transformed into the time domain using the IFFT block. IFFT inputs that
are not used (that is, that do not correspond to any of the minislots used by the transmission burst) are set to zero.
42
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
The transmitter converts the output of the IFFT from parallel to serial and performs cyclic prefix addition and
Windowing in the time domain.
7.4.2.7
Cyclic Prefix and Windowing
A segment at the end of the IFFT output is prepended to the IFFT output; this is referred to as the Cyclic Prefix (CP)
of the OFDM symbol. For windowing purposes, another segment at the start of the IFFT output is appended to the
end of the IFFT outputthe roll-off period (RP).
The addition of a cyclic prefix enables the receiver to overcome the effects of inter-symbol-interference and intercarrier-interference caused by micro-reflections in the channel. Windowing maximizes channel capacity by
sharpening the edges of the spectrum of the OFDM signal. Spectral edges occur at the two ends of the spectrum of
the OFDM symbol, as well as at the ends of internal exclusion bands.
These topics are discussed in detail in Section 7.5.8.
7.4.3
Time and Frequency Synchronization
CM upstream frequency and timing of transmissions is based on downstream tracking, and in the case of timing,
also based upon receiving and implementing timing adjustments from the CMTS. This section describes the CM
upstream transmission performance requirements on frequency and timing which are based upon tracking the
downstream input to the CM, and implementing and operating upon commands from the CMTS.
7.4.3.1
Channel Frequency Accuracy
To support OFDMA upstream using legacy downstream, the CM has to implement the assigned channel frequency
within 50 parts per million, over a temperature range of 0 to 40 degrees C, and up to five years from date of
manufacture.
The CM MUST implement the Offset Frequency Adjustment to effect a change in upstream subcarrier frequency
within 10 Hz of the commanded change.
The CM MUST provide upstream frequency accuracy, relative to the downstream reference, of 20 Hz in at least
Five Sigma of upstream transmissions, under all the downstream conditions detailed in Section 7.5.9.
7.4.3.2
Channel Timing Accuracy
For OFDMA upstream, regardless of what is used for the timing master, the range adjustment will be given units
that match the extended timestamp for DOCSIS 3.1 [DOCSIS MULPIv3.1].
Specifically, this timestamp has an integer portion of 10.24 MHz clocks. It also has an integer portion of counting
1/20ths duration of 10.24 MHz clock period (this integer portion counts up to 20), and then it has a 4 bit binary
fractional portion so the CMs timing resolution MUST be (1/10.24 MHz) x (1/20) x (1/16) = 305 ps.
The CMTS MUST be able to send timing adjustment commands with a resolution of 305 ps.
The CM MUST implement the OFDMA Timing Adjust to within +/- 10 ns.
7.4.3.3
Upstream Subcarrier Clock Frequency
The CM MUST lock the Upstream Subcarrier Clock Frequency (25 kHz or 50 kHz) to the 10.24 MHz Master Clock
via the Downstream Subcarrier Clock, subject to the carrier phase noise and frequency offset requirements in this
section. The subcarrier frequency is specified as locked to the 10.24 MHz reference and given with 25 kHz
resolution (25 kHz = 10.24 MHz x 10/4096).
Maximal subcarrier frequency offset depends on the downstream OFDM occupied bandwidth which determines the
number and general distribution of Continuous Pilots. All upstream subcarrier frequency specifications assume a
downstream input to the CM per Sections 7.5.9 and 7.5.3.
For OFDMA upstream channels, the CMs upstream subcarrier frequency offset for each subcarrier (relative to the
Master Clock reference) with downstream occupied bandwidth of 192 MHz MUST be less than 0.1 ppm for five
sigma of upstream OFDMA transmissions; at subcarrier frequency of 204 MHz this corresponds to less than 20 Hz.
10/29/13
CableLabs
43
CM-SP-PHYv3.1-I01-131029
For OFDMA upstream channels, the CMs upstream subcarrier frequency offset for each subcarrier (relative to the
Master Clock reference) with downstream occupied bandwidth of 24 MHz MUST be less than 1 ppm for five sigma
of upstream OFDMA transmissions.
7.4.4
Forward Error Correction
DOCSIS 3.1 uses three Quasi-Cyclic Low-Density Parity-Check codes (QC-LDPC) for the upstream transmission,
as depicted in Table 73.
Table 73 - Upstream Codeword Parameters
Code
Code rate
Codeword size in
bits (Ni)
Information bits
(Ki)
Parity bits (Pi)
Long code
Medium code
Short code
89% (8/9)
85% (28/33)
75% (3/4)
16200
5940
1120
14400
5040
840
1800
900
280
Before FEC encoding, the CM MUST first map the input byte stream into a bit-stream such that the LSB of the first
byte is the first bit of the bit-stream.
7.4.4.1
FEC Codeword Selection
The choice of codeword sizes to be used in any given burst is based on the grant in the MAP message. The grant
indicates which minislots are assigned to a given burst and which upstream profile is to be used. The CM and CMTS
use this information to determine the total number of bits in the grant which are available to be used for FEC
information or parity.
Codewords are filled and transmitted in the following order, with codeword shortening applied according to rules
defined in Section 7.4.4.3:
Full long codewords (if present)
Shortened long codeword (if present)
Full medium codewords (if present)
Shortened medium codeword (if present)
Full short codewords (if present)
Shortened short codeword (if present)
Zero-pad (if present)
The CM MUST follow the FEC codeword selection algorithm defined by Matlab code in Section 7.4.4.1.1 to
determine the exact number, type, and size of the codewords to be used, and in what order.
The CMTS MUST follow the FEC codeword selection algorithm defined by the Matlab code in Section 7.4.4.1.1 to
determine the exact number, type, and size of the codewords to be used, and in what order.
7.4.4.1.1
FEC codeword Selection Algorithm
The FEC codeword selection algorithm is given by:

%
%
%
%
%
%
The total number of bits in the grant is given by rgrant_size

set values for codeword sizes
total bits = size including parity
info bits = information bits only
thresholds - if more bits than threshold, shorten this cw instead of
using a smaller one
% short codeword
SHORT_TOTAL_BITS = 1120;
44
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
SHORT_INFO_BITS = 840;
SHORT_PARITY_BITS = SHORT_TOTAL_BITS - SHORT_INFO_BITS;
SHORT_TOTAL_THRESH_BITS = SHORT_PARITY_BITS + 1;
SHORT_MIN_INFO_BITS = SHORT_INFO_BITS / 2;
% medium codeword
MED_TOTAL_BITS = 5940;
MED_INFO_BITS = 5040;
MED_PARITY_BITS = MED_TOTAL_BITS - MED_INFO_BITS;
MED_TOTAL_THRESH_BITS = 3421;
% long codeword
LONG_TOTAL_BITS = 16200;
LONG_INFO_BITS = 14400;
LONG_PARITY_BITS = LONG_TOTAL_BITS - LONG_INFO_BITS;
LONG_TOTAL_THRESH_BITS = 11881;
% variable rgrant_size is input
% set rgrant_size to desired input value in workspace
% initialize output variables
rlong_cws = 0;
rshortened_long_cws = 0;
rmed_cws = 0;
rshortened_med_cws = 0;
rshort_cws = 0;
rshortened_short_cws = 0;
rother_shortened_cw_bits = 0;
rshortened_cw_bits = 0;
rpad_bits = 0;
% intermediate variable to track type of last full codeword
rlast_full_cw = '';
% now begin calculation
bits_remaining = rgrant_size;
% if we don't have enough space to make at least a min size shortened
% short cw, then this grant is nothing but pad bits.
% NOTE: in the case, the CM should ignore the grant and should not
% transmit any bits at all in the grant.
if rgrant_size < SHORT_MIN_INFO_BITS + SHORT_PARITY_BITS
rpad_bits = rgrant_size;
bits_remaining = 0;
end
% make as many full long cws as possible
while bits_remaining >= LONG_TOTAL_BITS
rlong_cws = rlong_cws + 1;
bits_remaining = bits_remaining - LONG_TOTAL_BITS;
rlast_full_cw = 'Long';
end
% if remaining bits can make a shortened long codeword, do so
if bits_remaining >= LONG_TOTAL_THRESH_BITS
rshortened_long_cws = 1;
rshortened_cw_bits = bits_remaining;
bits_remaining = 0;
end
% if not, make as many med cws as possible with remaining bits
while bits_remaining >= MED_TOTAL_BITS
rmed_cws = rmed_cws + 1;
bits_remaining = bits_remaining - MED_TOTAL_BITS;
10/29/13
CableLabs
45
CM-SP-PHYv3.1-I01-131029
rlast_full_cw = 'Medium';
end
% if remaining bits can make a shortened med cw, do so
if bits_remaining >= MED_TOTAL_THRESH_BITS
rshortened_med_cws = 1;
bits_remaining = 0;
end
% if not, make as many short cws as possible with remaining bits
while bits_remaining >= SHORT_TOTAL_BITS
rshort_cws = rshort_cws + 1;
bits_remaining = bits_remaining - SHORT_TOTAL_BITS;
rlast_full_cw = 'Short';
end
% if remaining bits can make a shortened short cw, do so
if bits_remaining >= SHORT_TOTAL_THRESH_BITS
rshortened_short_cws = 1;
%
%
%
%
at this point we are definitely making this cw; however, we need

at least SHORT_MIN_INFO_BITS to put in it. If we do not have
that many, we will have to borrow from the last full cw, making
it also a shortened cw.
if (bits_remaining - SHORT_PARITY_BITS) >= SHORT_MIN_INFO_BITS

% no need to borrow bits
bits_remaining = 0;
else
% identify type/size of last full cw, then borrow
% SHORT_MIN_INFO_BITS from it
switch rlast_full_cw
case 'Long'
% change last full cw to a shortened cw
rlong_cws = rlong_cws - 1;
rshortened_long_cws = rshortened_long_cws + 1;
% number of bits in that cw is reduced by
% SHORT_MIN_INFO_BITS
rother_shortened_cw_bits = LONG_TOTAL_BITS - ...
SHORT_MIN_INFO_BITS;
% put those bits plus bits_remaining into the last
% shortened cw
rshortened_cw_bits = SHORT_MIN_INFO_BITS + ...
bits_remaining;
bits_remaining = 0;
case 'Medium'
% same steps as for long, just substitute medium
rmed_cws = rmed_cws - 1;
rshortened_med_cws = rshortened_med_cws + 1;
rother_shortened_cw_bits = MED_TOTAL_BITS - ...
bits_remaining;
bits_remaining = 0;
case 'Short'
% again, same steps as for long - now substitute short
rshort_cws = rshort_cws - 1;
rshortened_short_cws = rshortened_short_cws + 1;
rother_shortened_cw_bits = SHORT_TOTAL_BITS - ...
46
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
bits_remaining;
bits_remaining = 0;
end
end
end
% any space left over at this point has to be filled with pad bits (it
% cannot fit any cws)
if bits_remaining > 0
rpad_bits = bits_remaining;
bits_remaining = 0;
end
Based on the algorithm above, the minimum grant size allowed is:
SHORT_MIN_INFO_BITS + SHORT_PARITY_BITS = SHORT_INFO_BITS / 2 + SHORT_PARITY_BITS =
420 + 280 bits = 700 bits. This grant is sufficient for 52 bytes of information.
The CM SHOULD NOT transmit in any grant smaller than the minimum allowed grant size specified above.
The FEC codeword selection algorithm follows the procedure below:
If there are enough bits in the grant to create a full long codeword, do so. Continue creating full long codewords
until there are not enough bits remaining.
If the number of bits remaining is greater than or equal to the minimum allowed size for a shortened long
codeword, create such a codeword and end the burst.
Otherwise, if there are enough bits remaining to create a full medium codeword, do so. Continue creating full
medium codewords until there are not enough bits remaining.
If the number of bits remaining is greater than or equal to the minimum allowed size for a shortened medium
codeword, create such a codeword and end the burst.
Otherwise, if there are enough bits remaining to create a full short codeword, do so. Continue creating full short
codewords until there are not enough bits remaining.
If there are enough bits remaining to create a shortened short codeword containing at least the minimum
allowed number of information bits, do so and end the burst.
Otherwise, if there are enough bits remaining to create a shortened short codeword with fewer than the
minimum allowed number of information bits, remove a number of bits equal to the minimum allowed number
of short codeword information bits from the last full codeword, changing it to a shortened codeword. Add this
number of bits to the bits remaining and create a shortened short codeword using these bits, and end the burst.
Otherwise, there are not enough bits remaining to create a shortened short codeword ( i.e., fewer bits than the
number required for one information bit plus the applicable number of parity bits). These bits will be padded
with zeros by the CM and will be ignored by the CMTS.
If a grant does not contain enough bits to create any codewords, the CM should not transmit in the grant.
The reverse calculation to determine the grant size required to hold the desired number of bits, number of codewords
and codeword sizes is given in Appendix IV.
7.4.4.2
FEC Encoding
All three LDPC encoders are systematic. Every encoder encodes N-M information bits i0 ,, i N M 1 into a
codeword c = (i0 ,, i N M 1 , p0 ,, p M 1 ) by adding m parity bits obtained so that Hc T = 0 , where H is an mn
parity check matrix. The parity-check matrix can be divided into blocks of L*L submatrices, where L represents the
submatrix size or lifting factor. The parity-check matrix in compact circulant form is represented by an m n block
matrix:
10/29/13
CableLabs
47
CM-SP-PHYv3.1-I01-131029
H 1,1
H
2,1
H = H 3,1

H m,1
H 1, 2
H 2, 2
H 1,3
H 2,3
H 3, 2
H 3,3
H m, 2
H m ,3
H 1,n
H 2,n
H 3,n

H m,n
Each submatrix Hi,j is an LL all-zero submatrix or a cyclic right-shifted identity submatrix. The last n m submatrix columns represent the parity portion of the matrix. In this specification, the L L sub-matrix Hi,j is
represented by a value in {-, 0,, L-1}, where a -' value represents an all-zero submatrix, and the remaining values
represent an LL identity submatrix cyclically right-shifted by the specified value. The code rate is (nm)/n and a
codeword length is N=n L bits.
48
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
The CM MUST employ the following matrix table for the long code rate:
Rate= 89% (16200, 14400) code, m=5 rows n=45 columns, L=360
93 271
83 26 208 245 200
274 115 329 338 124
134 355 175 24 253 242

-
293
-
184 70 247 14 22
253 273 90
175 331 17 86
69 64 342
187 94 26 87 302
7 285 54
151 311 320 339
337
88 139
-
238 81
137 212
191 323 22
307
165
47 76 73 150 349 139 331 118 345 27 294
157 195 357 81 194 1 159 56 72 126 277 156 32 111 175
245 294 240 84 76 342 345 174 269 329
352 26 108 10 298 123 139 117
336 49 202 359 342
295 148 48 91 62 100 232 146 200 135 12
179
214
224 106
232
145 279 97 106 160 143
306 224
206
218 104 40 197 73 229 63
273 177 245 98 355 178 176 147
21 331 313 349 34 97 187 38
280
29 106 334
-
235 52 170 58
270 72
221 208
-
257 0
The CM MUST employ the following matrix table for the medium code rate:
142 158 113 124 92 44 93 70 172 3
25 44 141 160 50 45 118 84
54 172 145 28 55 19 159 22 96 12 85
64 66 97
128 5 158 120 51 171 65 141
63 11 112 114 61 123 72 55 114 20 53 114 42 33
28 160 102 44
26
163
2 168 158
84 126 9 169 174 147 24 145
52 159 75 74 46 71 42 11 108 153
72
66 163 50 46 17 175
-
67 82
-
1 115 8 108
42 83
-
92
22
39 121 84 101 171

-
41 138
4 177 151 131 139 117 36 18
23
49 89 63 179 10 75 161
34 47
177 19
The CM MUST employ the following matrix table for the short code rate:
10/29/13
14 12
37 45 26 24
34
35
26
10 16 16 34
23
51
12 28 22 46
16 51
37
52 33
38
11 22 23 43
14
51 16 31 13 39 27 33
36
51
19
25 29 19 18 52
8
45 48
27 53 13
9
CableLabs
46 10
49 20
34 39
49
CM-SP-PHYv3.1-I01-131029
7.4.4.3
Shortening of LDPC Codewords
Shortening of LDPC codewords is useful in order to optimize FEC protection for the payload. If a shortened
codeword is required, the CM MUST construct it as follows:
1. Binary zeros are appended to a reduced number of information bits at the input of the encoder.
2. The encoder calculates the parity bits.
3. The appended binary zeros are removed from the transmitted shortened codeword.
4. If the last shortened codeword is less than half the number of information bits, then one half the number of
information bits from the previous codeword is inserted into the last shortened codeword.
7.4.5
Data Randomization
The CM MUST implement a randomizer in the upstream modulator (shown in Figure 74 where the 23-bit seed
value is programmable.
At the beginning of each burst, the register is cleared and the seed value is loaded. The CM MUST use the seed
value to calculate the scrambler bit which is combined in an XOR with the first bit of data of each burst.
The CM MUST configure the randomizer seed value in response to the Upstream Channel Descriptor provided by
the CMTS.
The CM MUST use x^23+x^18+1 for the data randomizer polynomial.
Seed load / Frame reset
Delay
Element
1
Delay
Element
2
Delay
Element
3
Delay
Element
18
Delay
Element
22
Delay
Element
23
XOR
Randomizer
data output
Randomizer
data input
XOR
Figure 74 - Upstream Data Randomizer
7.4.6
Time and Frequency Interleaving and De-interleaving
Upstream transmissions can be affected by burst noise that reduces the SNR of all the subcarriers of a few
successive OFDMA symbols. Upstream transmissions may also be impacted by ingress, i.e., relatively narrowband
interferers, that can last for several symbol periods and therefore reduce the SNR of a subset of subcarriers over an
50
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
entire OFDMA frame. The purpose of the interleaver is to distribute the affected subcarriers over a number of FEC
blocks, enabling the FEC at the receiver to correct the corrupt data.
Time and frequency interleaving in the upstream are applied together in the CM as a single operation and hence
referred to as upstream interleaving. Similarly, time and frequency de-interleaving are performed together as a single
operation in the CMTS, and hence referred to here as upstream de-interleaving.
The CM MUST apply interleaving to upstream OFDMA subcarriers. The interleaving is applied after the FEC
encoder while the bits are being mapped QAM subcarriers, and before the OFDMA IFFT operation.
The CM MUST exclude any zero-valued minislots from the interleaving process.
The CM MUST apply interleaving to a sequence of 24 minislots of an OFDMA frame of a specific grant as
described in this section.
The CMTS MUST apply de-interleaving which is the inverse of the CM interleaving function carried out by the
CM.
The maximum number of minislots over which interleaving is applied is equal to 24. If the number of minislots of a
specific grant in one OFDMA frame is less than or equal to 24, then interleaving is applied over all of these
minislots.
If the number of minislots in a specific grant in one OFDMA frame is more than 24, say NMS_Total, then the CM
MUST partition these NMS_Total minislots into (_ /24) blocks of minislots, as uniformly as possible,
without the number of minislots in any block exceeding 24, using the algorithm given in the flow diagram shown in
Figure 75.
_
_ =
24
_ = _
_ = _
=1
() = (_/_)
_ = _ ()
_ = _ 1
=+1
_ == 0
No
Yes
End
Figure 75 - Calculating Number of Minislots in Each Block for Upstream Interleaving
10/29/13
CableLabs
51
CM-SP-PHYv3.1-I01-131029
The described algorithm in Figure 75 yields the sequence:

{ (), = 1, 2, , _}
There are Blks_Total of blocks of minislots, and in each block there are NMS(i) minislots. For each block of NMS(i)
minislots the CM must apply interleaving as described in this section.
Minislot 4
Minislot 4
Minislot 3
Minislot 3
Exclusion zone
Minislot 2
Exclusion zone
Frequency
Minislot 1
Minislot 2
Minislot 1
Time
Interleaving /
de-interleaving
is applied over
this 2-D timefrequency block
Figure 76 - Illustrating Minislots of a Grant over which Interleaving is Performed
Figure 76 shows an example of a block of four minislots over which interleaving is applied. The horizontal axis is
time. Every vertical column constitutes a segment of an OFDMA symbol. The vertical axis is frequency. Each
horizontal line is a set of subcarriers at a specific frequency over several symbols.
In the illustration in Figure 76, there is an exclusion zone between minislots 1 and 2. There is also an exclusion
zone between minislots 2 and 3. All four minislots are merged to form a two-dimensional grant for the purpose of
interleaving and de-interleaving. In the CM, the interleaving is applied first and then the exclusion zones are
introduced in mapping of the interleaved subcarriers onto OFDMA symbols.
Interleaving and de-interleaving are two-dimensional operations in the time-frequency plane.
7.4.6.1
Time and Frequency Interleaving
The system block diagram for interleaving is illustrated in Figure 77.
52
CableLabs
10/29/13
FEC
Encoded
Bit
Store
CM-SP-PHYv3.1-I01-131029
Bits to
Constellation
Mapping
(L x K)
cells
Insert
Exclusion
Zones
OFDMA
Symbols
Addressing
Algorithm
Figure 77 - Sample Interleaver Block Diagram
The two-dimensional array is addressed by coordinate pair (, ). The horizontal dimension is K, which is the
number of OFDMA symbols in the frame. The vertical dimension is L, which is the total number of subcarriers in all
the minislots that make up the grant in the current frame. Each element in this two-dimensional array is a member of
the set:
{Data subcarrier, Complementary data subcarrier, Pilot}
All data subcarriers in a minislot will have the same QAM constellation. All complementary data subcarriers in a
minislot will also have the same QAM constellation, but this will be lower in order than that of the data subcarriers
in that minislot.
Furthermore, the QAM constellations of data and complementary pilots need not be the same for all minislots in the
grant.
Interleaving involves the following two stages.
Writing the subcarriers in the cells of the two dimensional array of size (L x K).
The CM MUST follow the algorithm given in this section for placing data subcarriers and complementary data
subcarriers in the cells of this two-dimensional array.
The CM MUST NOT place any data subcarriers or complementary pilots at locations corresponding to pilots
which are also part of this two-dimensional array.
Reading data subcarriers as well as pilots along vertical columns of the two-dimensional array, in the ascending
order of the time dimension coordinate , inserting exclusion zones, if any, and passing these segments of
OFDMA symbols to the IFFT processor.
Figure 76 is for illustration only. An implementation may not necessarily have a separate FEC Encoded bit store.
The FEC encoder output may be mapped directly into QAM subcarriers and placed in the cells of the (L x K) twodimensional array. In that way the two-dimensional array may form the output buffer for the FEC encoder.
The Address Generation and the Bit Mapping algorithms need to know:
a)
Values of K and L
b) Locations of pilots
c)
Locations of complementary pilots
10/29/13
CableLabs
53
CM-SP-PHYv3.1-I01-131029
d) QAM constellations for data subcarriers of all minislots of the grant in the frame
e)
QAM constellations for complementary pilots of all minislots of the grant in the frame
f)
Minislot boundaries along the frequency dimension of the (L x K) array
The interleaving algorithm follows the flow diagram in Figure 78.
54
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
New Interleaving Block
Get next coordinate pair (, )

Using the Interleaver Address Generation Algorithm
Identify the minislot from coordinate ; from this identify

a) Pilot locations
b) Complementary pilot locations
c) QAM constellation of data subcarriers
d) QAM constellation of complementary pilots
Is (, ) a pilot
location?
Yes
No
Yes
Is (, ) a comp.
location?
No
Map FEC encoded bits into a

data subcarrier and store this in
location (, )
Map FEC encoded bits into a

complementary pilot and store
this in location (, )
Complete (L x K)
Block Written?
No
Yes
Read the memory along columns, i.e., along dimension , in the
ascending order of coordinate . Insert exclusion bands, if any, in
each column and send to IFFT processing to generate Scrambler
OFDMA symbols.
Figure 78 - Interleaving a Grant within an OFDMA Frame
10/29/13
CableLabs
55
CM-SP-PHYv3.1-I01-131029
The address generation algorithm for getting the next coordinate pair (, ) is described below. This makes use of
three bit-reverse counters.
1) Count_t
2) Count_f
3) Count_diagonal
The third counter is used to count the diagonals. This is because subcarriers are written in the two-dimensional t-f
array along diagonals. To write along diagonals in natural order, both the counters Count_t and Count_f have to be
incremented at the same time. Once one diagonal is written, the third counter Count_diagonal is incremented by
one.
However, in order to maximize the separation of successive subcarriers in the time-frequency plane, bit-reversed
counting is used in all of the above three counters. This ensures that successive subcarriers are not written on
successive locations in the diagonals.
The algorithm for generating the sequence of addresses (, f) is described below with sample C code given in
Appendix III.
Initialize three counters, Count_t, Count_f and Count_diagonal, to zero.
For each value of OFDM symbol index idx_t going from 0 to (K-1), implement the following 4 steps:
1)
For each value subcarrier index idx_f going from 0 to (L-1) implement the following 4 sub-steps:
a) Generate the component t of (t, f) by passing Count_t and parameter K to the Bit-Reverse counter
defined in the flow diagram of Figure 79.This returns t and a new counter value for Count_t.
b) Generate the component f of (t, f) by passing Count_f and parameter L to the Bit-Reverse counter
defined in the flow diagram of Figure 79. This returns f and a new counter value for Count_f.
c)
Increment Count_t by one modulo K1
d) Increment Count_f by one modulo L1

2)
3)
4)
Increment Count_diagonal by one modulo K1

Pass Count_diagonal and parameter K to the Bit-Reverse counter defined in the flow diagram of
Figure 79. This returns a new counter value for Count_diagonal.
Set Count_t to the value of Count_diagonal. Set Count_f to zero. Return to step 1.
The pseudo code given above will generate the entire sequence of addresses. This is for illustration purposes only. In
the actual implementation, the code may be modified to generate one address at a time, so that data may be saved in
the memory in parallel with address generation.
The above pseudo code segment contains a call to the function called Bit_Reverse_Count. The algorithm
implemented by this function is explained below with reference to the flow diagram of Figure 79. With no loss of
generality, this explanation uses the function call for Count_t.
The parameter K1 is defined as the smallest power of 2 that is equal to or greater than K. The minimum number of
binary bits needed to represent K is k1. Then, 1 = 21 . Similarly, parameter L1 is defined as the smallest power of 2
greater than L.
56
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
_
Reverse Bits making MSB the LSB and
the LSB the MSB to give number B
No
Yes
_ (_ + 1) 1
Equivalent to a k1-bit binary counter with overflow
bits discarded
Return with new counter value

_ and its bit-versed
version B
Figure 79 - Bit-Reversed Counter Implementation
Bit-reverse counting employs a modulo 21 counter. This is equivalent to a k1-bit counter with overflow bits
discarded. In bit-reversed counting the above counter is incremented beginning from its current value until the bitreversed version of the counter value is in the range [0, (K-1)].
The term bit-reversion is defined below. Let A be the value of Count_t and let B be its bit-reversed value. Let the
binary representation of A be given by:
1 1
= =0
2
Then B is given by:
7.4.7
1 1
1 1
= =0
2 = =0
21 1
Mapping of Bits to Cell Words
CMs are granted transmission opportunities by minislots, and minislots are associated with subcarriers. All
subcarriers of a specific type (data subcarriers, pilots, complementary pilots, null subcarriers) within a minislot have
the same modulation order, although different minislots may have different modulation orders; the modulation order
to be used is determined by the Profile associated with the minislot.
The CM MUST modulate the incoming serial binary bitstream from the data scrambler to constellation symbols
using the constellation mapping described in Section 7.4.8.2.
The CM MUST encode the bit stream such that the first bit is the least-significant bit of the constellation symbol
when bits are mapped into constellation symbols.
The CM MUST have the same nominal average power for all constellations.
10/29/13
CableLabs
57
CM-SP-PHYv3.1-I01-131029
7.4.8
Mapping and De-mapping Bits to/from QAM subcarriers
CMs are granted transmission opportunities by minislots, and minislots are associated with subcarriers. All
subcarriers of a specific type ( i.e., data subcarriers, pilots, complementary pilots or zero-valued subcarriers) within a
minislot have the same modulation order, although different minislots may have different modulation orders; the
modulation order to be used is determined by the Profile associated with the minislot.
7.4.8.1
Modulation Formats
The CM modulator MUST support zero valued subcarriers of upstream OFDMA channels.
The CM modulator MUST support BPSK, QPSK, 8-QAM, 16-QAM, 32-QAM, 64-QAM, 128-QAM, 256-QAM,
512-QAM, 1024-QAM, 2048-QAM, and 4096-QAM for subcarriers of upstream OFDMA channels.
The CMTS demodulator MUST support zero valued subcarriers of upstream OFDMA channels.
The CMTS demodulator MUST support BPSK, QPSK, 8-QAM, 16-QAM, 32-QAM, 64-QAM, 128-QAM, 256QAM, 512-QAM, and 1024-QAM for subcarriers of upstream OFDMA channels.
The CMTS demodulator SHOULD support 2048-QAM and 4096-QAM for subcarriers of upstream OFDMA
channels.
7.4.8.2
Constellation Mapping
The CM MUST encode the bit stream such that the first bit is the least-significant bit of the first QAM subcarrier
constellation m-tuple.
Time
0 1 1 +1 21 2
MSB
LSB
QAM m-tuple 1:
1 2 0
21 22
QAM m-tuple 2:
31 32 2
QAM m-tuple 3:
Figure 710 - Bitstream to QAM M-Tuple Mapping
The CM MUST modulate the interleaved m-tuples onto subcarriers using QAM constellation mappings described in
Annex A.
The CM MUST ensure that subcarriers of all QAM constellations have the same nominal average power using the
scaling factors given in Table A1 of Annex A.
The CMTS receiver MUST demodulate the incoming QAM constellation subcarriers of a minislot according to the
Profile associated with the minislot, with the first demapped value associated with the least-significant bit of the
constellation symbol.
7.4.9
REQ Messages
REQ messages are short messages used by the CM to request transmission opportunities from the CMTS. These
messages have a different structure than the data messages: they are always 56 bits long, they always use QPSK
modulation, do not apply any FEC and are block interleaved.
REQ message processing is described in Figure 711.
58
CableLabs
10/29/13
REQ
message
from MAC
CM-SP-PHYv3.1-I01-131029
Randomizer
Bits for
QPSK
symbol
Block
Interleaver
Pre-EQ
IFFT
CP and
Window
Pilot
Pattern
Figure 711 - REQ Messages Processing
The CM MUST randomize REQ messages using the randomizer described in Section 7.4.5.
The CM MUST modulate REQ messages using QPSK.
The CM MUST use the subslot minislots with the pilot patterns as specified in Sections 7.4.16.4 and 7.4.16.5 for 25
KHz and 50 kHz subcarrier spacing.
The CM MUST write the REQ messages QPSK symbols into sublots as described in Section 8.2.3.2.
The CM MUST use the same CP size and RP size used for the data transmission.
7.4.10 IDFT
The upstream OFDMA signal transmitted by the CM is described using the following IDFT equation:
2
2 ) , = 0, 1, , ( 1)
() =
() exp(
=0
1
Where N equals 2048 with 50 KHz subcarrier spacing and 4096 with 25 KHz subcarrier spacing. The resulting time
domain discrete signal, x(i), is a baseband complex-valued signal, sampled at 102.4 Msamples per second.
In this definition of the IDFT X(0) is the lowest frequency component; and X(N-1) is the highest frequency
component.
The IDFT operation is illustrated in Figure 712.
10/29/13
CableLabs
59
CM-SP-PHYv3.1-I01-131029
Frequency
()
k=0
k = N-1

Transform
Time
()
i=0
i = N-1
Figure 712 - Inverse Discrete Fourier Transform

Cyclic prefix and windowing are applied in the upstream transmission. Cyclic prefix is added in order to enable the
receiver to overcome the effects of inter-symbol interference (ISI) and inter-carrier interference caused by microreflections in the channel. Windowing is applied in order to maximize channel capacity by sharpening the edges of
the spectrum of the OFDMA signal. Spectral edges occur at the two ends of the spectrum of the OFDM symbol, as
well as at the ends of internal exclusion bands.
In the presence of a micro-reflection in the transmission medium, the received signal is the sum of the main signal
and the delayed and attenuated micro-reflection. As long as this delay () is less than the time duration of the cyclic
prefix (TCP), the CMTS receiver can trigger the FFT to avoid any inter-symbol or inter-carrier interference due to
this micro reflection, as shown in Figure 713.
Cyclic
Prefix
OFDM Symbol
Cyclic
Prefix
Cyclic
OFDM Symbol
Cyclic
OFDM Symbol
OFDM Symbol
FFT trigger-point region

Figure 713 - Signal with Micro-Reflection at the Receiver
60
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
If the delay of the micro-reflection exceeds the length of the cyclic prefix, the ISI resulting from this microreflection is:
=
where:
( )2
is the delay introduced by the micro-reflection
is the cyclic prefix length in s
is the relative amplitude of the micro-reflection
is FFT duration (20 or 40 s)
The inter-carrier-interference introduced by this micro-reflection is of the same order as the ISI.
The CM transmitter MUST apply the configured CP and Windowing as described in Section 7.4.11.1.
The CM transmitter MUST support the cyclic prefix values defined in Table 74.
The CMTS receiver MUST support the cyclic prefix values defined in Table 74.
Table 74 - Cyclic Prefix (CP) Values
Cyclic Prefix (s)
Cyclic Prefix Samples (Ncp)
0.9375
1.25
1.5625
1.875
2.1875
2.5
2.8125
3.125
3.75
5.0
6.25
96
128
160
192
224
256
288
320
384
512
640
In Table 74 the cyclic prefix (in s) is converted into samples using the sample rate of 102.4 Msamples/s.
Windowing is applied in the time domain by tapering (or rolling off) the edges using a raised cosine function.
The CMTS MUST support the eight roll-off prefix values listed in Table 75.
The CM MUST support the eight roll-off prefix values listed in Table 75.
The CMTS MUST NOT allow a configuration in which the Roll-Off Period value is the Cyclic Prefix value.
Table 75 - Roll-Off Prefix (RP) Values
10/29/13
Roll-Off Period (s)
Roll-Off Period Samples (NRP)
0
0.3125
0.625
0.9375
1.25
1.5625
1.875
2.1875
0
32
64
96
128
160
192
224
CableLabs
61
CM-SP-PHYv3.1-I01-131029
The Roll-Off Period is given in s and in number of samples using the sample rate of 102.4 Msamples/s.
7.4.11.1 Cyclic Prefix and Windowing Algorithm
The algorithm for cyclic prefix extension and windowing is described here with reference to Figure 714.
The CM MUST support cyclic prefix extension and windowing as described in this section.
62
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Time
N-point IDFT Output
N-point IDFT Output

NCP
N-point IDFT Output

NCP
NRP
N-point IDFT Output

NCP
NRP
NRP
NRP
+
N + NCP
Figure 714 - Cyclic Prefix and Windowing Algorithm
10/29/13
CableLabs
63
CM-SP-PHYv3.1-I01-131029
Processing begins with the N-point output of the IDFT. Let this be:
{ (0), (1), , ( 1)}
The NCP samples at the end of this N-point IDFT are copied and prepended to the beginning of the IDFT output to
give a sequence of length (N+NCP):
{( ), ( + 1), , ( 1), (0), (1), , ( 1)}
The NRP samples at the start of this N-point IDFT are copied and appended to the end of the IDFT output to give a
sequence of length (N+NCP+NRP):
{( ), ( + 1), , ( 1), (0), (1), , ( 1), (0), (1), . .. , ( 1)}
Let this extended sequence of length (N+NCP+NRP) be defined as:

{(), = 0, 1, , ( + + 1)}
NRP samples at both ends of this extended sequence are subject to tapering. This tapering is achieved using a raisedcosine window function; a window is defined to be applied to this entire extended sequence. This window has a flat
top and raised-cosine tapering at the edges, as shown in Figure 715.
N + NCP
NRP/2
Figure 715 - Tapering Window
The window function w(i) is symmetric at the center; therefore, only the right half of the window is defined in the
following equation:
+
+ +
+ = 1.0 , = 0, 1, ,
1
2
2
+
+ +
=
1 sin

+ 1/2,
)
2
2
2
( +
+ +
+
, ,
1
2
2
for =
Here,
=
defines the window function for ( + + )/2 samples. The complete window function of length
(N + NCP + NRP) is defined using the symmetry property as:
+ +
+ +
1 =
+ ,
2
2
+ +
1
for = 0, 1, ,
2
64
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
This yields a window function (or sequence): {(), = 0, 1, , ( + + 1)}. The length of this
sequence is an even-valued integer.
The above window function is applied to the sequence {()}:
() = () (), = 0,
1, , ( + + 1)
Each successive set of N samples at the output of the IDFT yields a sequence () of length ( + + ). Each
of these sequences is overlapped at each edge by NRP samples with the preceding and following sequences, as shown
in the last stage of Figure 714. Overlapping regions are added together.
To define this "overlap and add" function mathematically, consider two successive sequences 1 () and 2 (). The
overlap and addition operations of these sequences are defined using the following equation:
1 ( + + ) + 2 (), for = 0, 1, , 1
That is, the last samples of sequence 1 () are overlapped and added to the first samples of sequence 2 ().
7.4.11.2 Parameter Optimization
7.4.11.2.1 Impacts of Cyclic Prefix and Windowing

The combination of cyclic prefix insertion and windowing can impact OFDM symbol duration: once the CP and RP
additions have been made, the length of the extended OFDM symbol is (N + NCP + NRP) samples. Of this, (NRP/2)
samples are within the preceding symbol, and (NRP/2) samples are within the following symbol. This yields a
symbol period of (N + NCP) samples.
In addition, successive symbols interfere with each other by (NRP/2) samples. Therefore, the non-overlapping flat
segment of the transmitted symbol = (N + NCP - NRP).
There are eleven possible values for NCP and eight possible values for NRP. This gives 88 possible values for .
However, combinations NRP NCP are not permitted. This limits the number of possible combinations for .
The user would normally select the cyclic prefix length NCP to meet a given delay spread requirement in the channel.
Then the user would select the NRP to meet the roll-off (i.e., the parameter) requirement. Increasing parameter
leads to sharper spectral edges in the frequency domain.
However, increasing NRP for a given NCP reduces the non-overlapping flat region of the symbol, thereby reducing
the ability of the receiver to overcome inter-symbol-interference. Similarly, increasing NCP for a given NRP does
reduce the roll-off parameter .
7.4.11.2.2 Joint Optimization of Cyclic Prefix and Windowing Parameters
It is clear from the section above that the parameters and have to be jointly optimized for given channel,
taking into account the following properties of the channel:
a)
Bandwidth of the transmitted signal
b)
Number of exclusion zones in the transmitted bandwidth
c)
Channel micro-reflection profile
d)
QAM constellation
The QAM constellation defines the tolerable inter-symbol and inter-carrier interference. This in turn defines the
cyclic prefix for a given micro-reflection profile. The bandwidth of the transmitted signal and the number of
exclusion zones define the sharpness of the spectral edges, and hence the amount of tapering. However, the amount
of tapering and the flat region of the cyclic prefix are not independent variables. Therefore, an optimization program
is needed to identify optimum values for and for the above parameters. This optimization is important
because it does have significant impact on channel capacity, i.e., the bit rate.
The joint optimization of and is left to the network operator.
10/29/13
CableLabs
65
CM-SP-PHYv3.1-I01-131029

A DOCSIS 3.1 CM is required to generate up to 8 channels of legacy DOCSIS plus up to 2 channel-blocks of
OFDMA as defined in Sections 6.2.1 and 7.2.1.
A CMs Transmit Channel Set (TCS) is the combination of legacy channels and OFDMA channels being
transmitted by the CM.
With N legacy channels in its TCS, and BWOFDMA being the combined occupied bandwidth of the OFDMA
channel(s) in its TCS, the CM is said to have Neq = 4*N + ceil(BWOFDMA (MHz)/1.6 MHz) "equivalent DOCSIS
channels" in its TCS.
"Equivalent channel power" of a legacy DOCSIS channel refers to the power in 1.6 MHz of spectrum assuming the
channel has occupied bandwidth of 6.4 MHz (5.12 MHz modulation rate).
The "equivalent channel power" of an OFDMA channel is the average power of the OFDMA subcarriers of the
channel normalized to 1.6 MHz bandwidth. This equivalent channel power of an OFDMA channel is denoted as
PSD1.6.The TCS has N legacy (N from zero to eight) plus zero, one, or two OFDMA channels, but also is described
as having Neq number of equivalent DOCSIS channels.
Each channel in the TCS is described by its reported power, which is the power of the channel when it is fully
granted. Each channel is also characterized by its "equivalent channel power," which is the channel power
normalized to 1.6 MHz (Power Spectral Density of the average power of the channel multiplied by 1.6 MHz).
For example, legacy DOCSIS channels have equivalent channel power of Channel Power 6.02 dB, since the
equivalent channel power is calculated for 1.6 MHz bandwidth but the calculation (for DOCSIS 3.1) also sets the
channel bandwidth at 6.4 MHz regardless of actual modulation rate.
7.4.12.1 Transmit Power Requirements
The transmit power requirements are a function of the number and occupied bandwidth of the OFDMA channels and
the number of legacy channels in the TCS. The minimum highest value of the total power output of the CM is 65
dBmV, although higher values are allowed. The total maximum power is distributed among the channels in the TCS,
based on equal power spectral density when the OFDMA channels are fully granted to the CM. This ensures that
each channel can be set to a power range (within the DRW) between its maximum power, Phi, and minimum power,
Plow, and that any possible transmit grant combination can be accommodated without exceeding the transmit power
capability of the CM.
Maximum equivalent channel power (PSD1.6Max) is calculated as PSD1.6Max = 65 dBmV 10log10(Neq).
The CMTS MUST limit the commanded PSD1.6Max to 53.2 dBmV if Neq is 15. This enforces a maximum power
spectral density of 65 dBmV per 24 MHz.
Legacy channels are allocated a maximum power of PSD1.6Max + 6 dB because they are considered equivalent to 4
equivalent DOCSIS channels.
The minimum value allowed is PSD1.6Min = 17 dBmV.
The CM MUST NOT transmit on any channel in the TCS at a value higher than Phi or lower than Plow. The values
for Phi and Plow are functions of PSD1.6Max, PSD1.6Min, and Neq.
The CMTS sends transmit power level commands and pre-equalizer coefficients to the CM [DOCSIS MULPIv3.1]
to compensate for upstream plant conditions. The top edge of the DRW is set to a level, Pload_min_set, close to the
highest PSD1.6 transmit channel to optimally load the DAC. In extreme tilt conditions, some of the channels will be
sent commands to transmit at lower PSD1.6 values that use up a significant portion of the DRW. Additionally, the
pre-equalizer coefficients of the OFDMA channels will also compensate for plant tilts. The CMTS enforces a DRW
of 12 dB [DOCSIS MULPIv3.1] which is sufficient to accommodate plant tilts of up to 10 dB from lower to upper
edge of the upstream band. Since the fidelity requirements are specified in flat frequency conditions from the top of
the dynamic range window, it is desirable to maintain CM transmission power levels as close to the top of the
dynamic range window as possible. When conditions change sufficiently to warrant it, a global reconfiguration time
should be granted and the top of the DRW adjusted to maintain the best transmission fidelity and optimize system
performance.
66
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
7.4.12.2 Global Reconfiguration Time for OFDMA Channels

"Global reconfiguration time" is defined as the inactive time interval provided between active upstream
transmissions, which simultaneously satisfies the requirement in Section 6.2.20 for all TDMA channels in the TCS
and the requirement in Section 6.2.20 for all S-CDMA channels in the TCS and the requirement in Section 7.4.12for
OFDMA.
With Multiple Transmit Channel mode enabled, the CMTS SHOULD provide global reconfiguration time to a CM
before (or concurrently as) the CM has been commanded to change any upstream channel transmit power by 3 dB
cumulative since its last global reconfiguration time, excluding transmit power changes due to UCD-induced change
in Phi.
7.4.12.3 OFDMA Fidelity Requirements
The following requirements assume that any pre-equalization is disabled, unless otherwise noted.
When channels in the TCS are commanded to the same equivalent powers, the reference signal power in the "dBc"
definition is to be interpreted as the measured total grant power. When channels in the TCS are commanded to
different equivalent channel powers, the commanded total power of the transmission is computed, and a difference
is derived compared to the commanded total power which would occur if all channels had the same PSD1.6 as the
highest equivalent channel power in the TCS, whether or not the channel with the largest equivalent channel power
is included in the grant. Then this difference is added to the measured total transmit power to form the reference
signal power for the "dBc" spurious emissions requirements.
7.4.12.3.1 Spurious Emissions
The noise and spurious power generated by the CM MUST NOT exceed the levels given in Table 76, Table 77,
and Table 78. Up to five discrete spurs can be excluded from the emissions requirements listed in Table 76, Table
77 and Table 78 and have to be less than -42 dBc relative to a single subcarrier power level.
SpurFloor is defined as:
SpurFloor = max{ -57 + 10*log10(100% Grant Spectrum/192 MHZ), -60} dBc
Under-grant Hold Number of Users is defined as:
Under-grant Hold Number of Users = Floor{ 0.2 + 10^( (-44 - SpurFloor)/10) }
Under-grant Hold Bandwidth is defined as:
Under-grant Hold Bandwidth = (100% Grant Spectrum)/(Under-grant Hold Number of Users)
When Multiple Transmit Channel mode is enabled, these spurious performance requirements only apply when the
CM is operating within certain ranges of values for Pload_n, for n = 1 to the number of upstream channels in the TCS,
and for granted bandwidth of Under-grant Hold Bandwidth or larger; where Pload_1 the highest loaded channel in this
specification (i.e., its power is the one closest to Phi).
When a modem is transmitting over a bandwidth of less than Under-grant Hold Bandwidth, the spurious emissions
requirement limit is the power value (in dBmV), corresponding to the specifications for the power level associated
with a grant of bandwidth equal to Under-grant Hold Bandwidth.
When Multiple Transmit Channel mode is enabled and there are two or more channels in the TCS, the CMs
spurious performance requirements MUST be met only when the equivalent DOCSIS channel powers (PSD1.6) are
within 6 dB of Pload_min_set (Pload_min_set +6>=Pload_n>= Pload_min_set).
Further, the CMs spurious emissions requirements MUST be met with two or more channels in the TCS only when
Pload_1 = Pload_min_set. When Pload_1 < Pload_min_set, the spurious emissions requirements in absolute terms are relaxed by
Pload_1 - Pload_min_set.
When a modem is transmitting with any equivalent DOCSIS channel power with loading Pload_min_set + 16 dB or
lower (Pload> Pload_min_set+ 16 dB), the spurious emissions requirement limits are not enforced.
With Multiple Transmit Channel mode enabled, the spurious performance requirements do not apply to any
upstream channel from the time the output power on any active upstream channel has varied by more than 3 dB
10/29/13
CableLabs
67
CM-SP-PHYv3.1-I01-131029
since the last global reconfiguration time through the end of the next global reconfiguration time changes, excluding
transmit power changes due to UCD-induced change in Phi [DOCSIS MULPIv3.1].
In Table 76, inband spurious emissions includes noise, carrier leakage, clock lines, synthesizer spurious products,
and other undesired transmitter products. It does not include ISI. The measurement bandwidth for inband spurious
for OFDM is equal to the Subcarrier Clock Frequency (25 kHz or 50 kHz) and is not a synchronous measurement.
The signal reference power for OFDMA inband spurious emissions is the total transmit power measured and
adjusted (if applicable) as described in Section 7.4.12.3, and then apportioned to a single data subcarrier.
For S-CDMA and TDMA, the measurement bandwidth is the modulation rate (e.g., 1280 to 5120 kHz), and the
requirement is -50 dBc. All requirements expressed in dBc are relative to the largest equivalent DOCSIS channel
power in the TCS, whether it is being transmitted or not.
The measurement bandwidth is 160 kHz for the Between Bursts (none of the channels in the TCS is bursting) specs
of Table 76, except where called out as 4 MHz or 250 kHz. The signal reference power for Between Bursts
transmissions is the total transmit power measured and adjusted (if applicable) as described in Section 7.4.12.3.
The Transmitting Burst specs apply during the mini-slots granted to the CM (when the CM uses all or a portion of
the grant), and for 20 s before and after the granted mini-slot for OFDMA. The Between Bursts specs apply except
during a used grant of mini-slots on any active channel for the CM, and 20 us before and after the used grant for
OFDMA. The signal reference power for Transmitting Burst transmissions, other than inband, is the total transmit
power measured and adjusted (if applicable) as described in Section 7.4.12.3.
For the purpose of spurious emissions definitions, a granted burst refers to a burst of minislots to be transmitted at
the same time from the same CM; these minislots are not necessarily contiguous in frequency.
Table 76 - Spurious Emissions
Parameter
Transmitting Burst
Between Bursts
Inband
-60 dBc OFDMA

-50 dBc SCDMA/TDMA
(NOTE: also see MER
requirement)
-72 dBc
Adjacent Band
See Table 78
-72 dBc
Within the upstream operating range 5-42

MHz or 5-85 MHz, or 5-204 MHz
(excluding assigned channel, adjacent
channels)
See Table 77
-72 dBc
For the case where the upstream operating

range is 5-42 MHz:
CM Integrated Spurious Emissions Limits
(all in 4 MHz, includes discretes)1
42 to 54 MHz
54 to 60 MHz
60 to 88 MHz
88 to 1218 MHz
-40 dBc
-35 dBmV
-40 dBmV
-45 dBmV
-26 dBmV
-40 dBmV
-40 dBmV
max(-45 dBmV, -40 dB ref downstream)2

range is 5-42 MHz:
CM Discrete Spurious Emissions Limits1
42 to 54 MHz
54 to 88 MHz
88 to 1218 MHz
-50 dBc
-50 dBmV
-50 dBmV
-36 dBmV
-50 dBmV
-50 dBmV
68
CableLabs
10/29/13
Parameter
CM-SP-PHYv3.1-I01-131029
Transmitting Burst
Between Bursts

range is 5-85 MHz:
(all in 4 MHz, includes discrete spurs)1
85 to 108 MHz
85 to 108 MHz (Should)
108 to 136 MHz
136 to 1218 MHz
-45 dBc
-50 dBc
-40 dBmV
-45 dBmV
-31 dBmV
-36 dBmV
-40 dBmV

range is 5-85 MHz:
85 to 108 MHz
108 to 1218
-50 dBc
-50 dBmV
-36 dBmV
-50 dBmV

range is 5-204 MHz:
(all in 4 MHz, includes discrete spurs)1
204 to 258 MHz
204 to 258 MHz (Should)
258 to 1218 MHz
-50 dBc
-60 dBc
-45 dBmV
-72 dBc
-72 dBc

range is 5-204 MHz:
204 to 258 MHz
258 to 1218 MHz
-50 dBc
-50 dBmV
-36 dBmV
-50 dBmV
Table Notes:
Note 1
These spec limits exclude a single discrete spur related to the tuned received channel; this single discrete spur must
be no greater than -40 dBmV.
Note 2
"dB ref downstream" is relative to the received downstream signal level. Some spurious outputs are proportional to the
receive signal level.
Note 3
Relative to the previous transmission
7.4.12.3.1.1 Spurious Emissions in the Upstream Frequency Range

Table 77 lists the required spurious level in a measurement interval. The initial measurement interval at which to
start measuring the spurious emissions (from the transmitted bursts modulation edge) is 400 kHz from the edge of
the transmissions modulation spectrum. Measurements should start at the initial distance and be repeated at
increasing distance from the carrier until the upstream band edge or spectrum adjacent to other modulated spectrum
is reached.
For OFDMA transmissions with non-zero transmit windowing, the CM MUST meet the required performance
measured within the 2.0 MHz adjacent to the modulated spectrum using slicer values from a CMTS burst receiver or
equivalent, synchronized to the downstream transmission provided to the CM.
In the rest of the spectrum, the CM MUST meet the required performance measured with a bandpass filter (e.g., an
unsynchronized measurement).
For OFDMA transmissions with zero transmit windowing, CM MUST meet the required performance using
synchronized measurements across the complete upstream spectrum.
For legacy transmissions, the measurement is performed in the indicated bandwidth and distance from the
transmitted legacy channel edge.
10/29/13
CableLabs
69
CM-SP-PHYv3.1-I01-131029
Spurious emissions allocation for far out spurious emissions =

Round{ SpurFloor + 10*log10( Measurement bandwidth/Under-grant hold Bandwidth),0.1}.
For transmission bursts with modulation spectrum less than the Under-grant Hold Bandwidth, the spurious power
requirement is calculated as above, but increased by 10*log10 (Under-grant Hold Bandwidth/Grant Bandwidth).
Table 77 - Spurious Emissions Requirements in the Upstream Frequency Range for Grants of
1
Under-grant Hold Bandwidth and Larger
100% Grant
Spectrum
(MHz)
SpurFloor (dBc)
Under-grant
Hold #Users
Under-grant Hold
Bandwidth (MHz)
Measurement
Bandwidth (MHz)
Specification in
the Interval (dBc)
Up to 64
-60.0
40
100% Grant
Spectrum/40
1.6
Round{ SpurFloor +
10*log10(
Measurement
Bandwidth/Undergrant Hold
Bandwidth),0.1}
[Ex: 22 MHz]
[0.55 MHz]
[-55.4]
[Ex: 46 MHz]
[1.15 MHz]
[-58.6]
Greater than 64,

up to 96
-60.0
40
100% Grant
Spectrum/40
[Ex 94 MHz]
3.2
Round{ SpurFloor +
10*log10(
Measurement
Bandwidth),0.1
[2.35 MHz]
Greater than 96,

up to 192
max{ -57 +
10*log10(100%
Grant
Spectrum/192
MHZ), -60}
Floor{ 0.2 + 10^(

(-44 SpurFloor)/10) }
100% Grant
Spectrum)/(Undergrant Hold Number of
Users)
[Ex: 142 MHz]
[-58.3]
[27]
[5.3]
[Ex: 190 MHz]
[-57.0]
[20]
[9.5
Greater than 192
max{ -57 +
10*log10(100%
Grant
Spectrum/192
MHZ), -60}
Floor{ 0.2 + 10^(

100% Grant
Spectrum)/(Undergrant Hold Number of
Users)
[Ex: 204 MHz]
[-56.8]
[19]
[10.5]
[-58.7]
9.6
Round{ SpurFloor +
10*log10(
Measurement
Bandwidth),0.1}
[-55.7]
[-57.0]
12.8
Round{ SpurFloor +
10*log10
(Measurement
Bandwidth),0.1}
[-55.9]
Note 1 Spurious Emissions Requirements in the Upstream Frequency Range Relative to the Per Channel Transmitted Burst
Power Level for Each Channel for Grants of Under-grant Hold Bandwidth and Larger.
Within the measurement bandwidth of Table 77, spurious emissions measured for individual subcarriers MUST
contain no more than +3 dB power larger than the required average power of the spurious emissions in the full
measurement bandwidth.
For legacy transmissions, and optionally for OFDMA transmissions, bandpass measurements rather than
synchronous measurements may be applied.
As an example illustrating the smaller measurement bandwidth requirements, consider 94 MHz 100% grant
spectrum, with -58.7 dBc spurious emissions allowed in 3.2 MHz measurement bandwidth, with the measurement
bandwidth starting as close as 400 kHz from the modulation edge of the transmitted burst. If the subcarrier spacing
69 is 25 kHz, there are 128 subcarriers in the 3.2 MHz measurement bandwidth. Each subcarrier has, on average, a
requirement of -58.7 dBc 21.1 dB = -79.8 dBc, but the requirement is relaxed to -79.8 dBc + 3 dB = -76.8 dBc
70
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
(noting that -21.07 dB corresponds to 1/128th). The under-grant hold bandwidth is 2.35 MHz for this example. When
a 100% grant has 65 dBmV transmit power, a grant of 2.4 MHz has 49.1 dBmV power. With a single OFDMA
channel and its 100% grant power at 65 dBmV, the spurious emissions requirement with a grant of 2.4 MHz,
measured in 25 kHz is 49.1 dBmV 76.8 dBc = -27.7 dBmV. -76.8 dBc corresponds to -57.1 dBr for this example
(since 2.35 MHz/25 kHz is a factor of 94, or 19.7 dB).
7.4.12.3.1.2 Adjacent Channel Spurious Emissions
Spurious emissions from a transmitted burst may occur in adjacent spectrum, which could be occupied by a legacy
carrier of any allowed modulation rate or by OFDMA subcarriers.
Table 78 lists the required adjacent channel spurious emission levels when there is a transmitted burst with
bandwidth at the Under-grant Hold Bandwidth. The measurement is performed in an adjacent channel interval of
400 kHz adjacent to the transmitted burst modulation spectrum. For OFDMA transmissions, the measurement is
performed starting on an adjacent subcarrier of the transmitted spectrum (both above and below), using the slicer
values from a CMTS burst receiver or equivalent synchronized to the downstream transmission provided to the CM.
For legacy transmissions, the measurement is performed in an adjacent channel interval of 400 kHz bandwidth
adjacent to the transmitted legacy channel edge.
Firstly, it should be noted that the measurement bandwidth for Table 78 is less than the measurement bandwidths in
Table 77. Thus comparing the two tables in terms of the specification "dBc" values requires appropriate scaling.
Secondly, Table 78 provides specification "dBc" only for grants of a specific amount for each row, while Table 7
7 provides "dBc" specification for grants of all sizes from the Under-grant Hold Bandwidth to 100%.
For transmission bursts with modulation spectrum less than the Under-grant Hold Bandwidth, the spurious power
requirement is calculated as above, but increased by 10*log10(Under-grant Hold Bandwidth/Grant Bandwidth).
For transmission bursts with modulation spectrum greater than the Under-grant Hold Bandwidth, the spurious power
requirement in the adjacent 400 kHz is calculated by converting the requirement to absolute power "dBmV" for a
grant of precisely Under-grant Hold Bandwidth from Table 78, and similarly computing the absolute power
"dBmV" from Table 77 for a grant equal to:
The Given Grant - The Under-grant Hold Bandwidth.
Then the absolute power calculated from Table 77 is scaled back in exact proportion of 400 kHz compared to the
measurement bandwidth in Table 77. Then the power from Table 78 is added to the scaled apportioned power
from Table 77 to produce the requirement for the adjacent 400 kHz measurement with a larger grant than the
Under-grant Hold Bandwidth. The requirement for adjacent spurious power in adjacent 400 kHz is:
P1(Grant Bandwidth Under-grant Hold Bandwidth) = absolute power derived from Table 77
(dBmV)
P2(Under-grant Hold Bandwidth) = absolute power derived from Table 78
(dBmV)
P1scaled = P1 * (0.4 MHz)/(Measurement Bandwidth (MHz) used in Table 77)
(dBmV)
Pspec_limit = P1scaled + P2
(dBmV)
Within the measurement bandwidth of Table 78, spurious emissions measured for individual subcarriers MUST
contain no more than +3 dB power larger than the required average power of the spurious emissions in the full
measurement bandwidth. For legacy transmissions, and optionally for OFDMA transmissions, bandpass
measurements rather than synchronous measurements may be applied.
10/29/13
CableLabs
71
CM-SP-PHYv3.1-I01-131029
Table 78 - Adjacent Channel Spurious Emissions Requirements Relative to the Per Channel
Transmitted Burst Power Level for Each Channel
100% Grant
Spectrum
(MHz)
Up to 64
SpurFloor (dBc)
Under-grant
Hold #Users
Under-grant Hold
Bandwidth (MHz)
-60.0
40
100% Grant
Spectrum/40
[Ex: 22 MHz]
Specification in
Adjacent 400 kHz
With Grant of Undergrant Hold Bandwidth
(dBc)
Round{10*log10(
((10^(SpurFloor/10)) +
(10^(-57/10))) x(0.4
MHz/Under-grant Hold
Bandwidth)),0.1}
[-56.6]
[1.15 MHz]
[-59.8]
-60.0
40
100% Grant
Spectrum/40
max{ -57 +
10*log10(100%
Grant
Spectrum/192
MHZ), -60}
Floor{ 0.2 + 10^(

100% Grant
Spectrum)/Undergrant Hold Number of
Users
[Ex 94 MHz]
Greater than 96
0.4 MHz
[0.55 MHz]
[Ex: 46 MHz]
Greater than 64,
up to 96
Measurement
Bandwidth
(MHz)
0.4 MHz
Round{10*log10(
(10^(-57/10))) x(0.4
Bandwidth)),0.1}
0.4 MHz
Round{10*log10(
(10^(-57/10))) x(0.4
Bandwidth)),0.1}
[2.35 MHz]
[-62.9]
Round nearest 0.1

dB
[Ex: 142 MHz]
[-58.3]
[27]
[5.3]
[-65.8]
[Ex: 190 MHz]
[-57.0]
[20]
[9.5]
[-67.7]
[Ex: 204 MHz]
[-56.8]
[19]
[10.5]
[-68.1]
7.4.12.3.2 Spurious Emissions During Burst On/Off Transients

The CM MUST control spurious emissions, prior to and during ramp-up, during and following ramp-down, and
before and after a burst.
The CMs on/off spurious emissions, such as the change in voltage at the upstream transmitter output, due to
enabling or disabling transmission, MUST be no more than 50 mV.
The CMs voltage step MUST be dissipated no faster than 4 s of constant slewing. This requirement applies when
the CM is transmitting at +55 dBmV or more per channel on any channel.
At backed-off transmit levels, the CMs maximum change in voltage MUST decrease by a factor of 2 for each 6 dB
decrease of power level in the highest power active channel, from +55 dBmV per channel, down to a maximum
change of 3.5 mV at 31 dBmV per channel and below. This requirement does not apply to CM power-on and poweroff transients.
7.4.12.3.3 MER Requirements
Transmit modulation error ratio (TxMER or just MER) measures the cluster variance caused by the CM during
upstream transmission due to transmitter imperfections. The terms "equalized MER" and "unequalized MER" refer
to a measurement with linear distortions equalized or not equalized, respectively, by the CM receive equalizer. The
requirements in this section refer only to unequalized MER, as described for each requirement. MER is measured on
each modulated data subcarrier and non-boosted pilot (MER is computed based on the unboosted pilot power) in a
mini-slot of a granted burst and averaged for all the subcarriers in each mini-slot. MER includes the effects of Inter-
72
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Carrier Interference (ICI), spurious emissions, phase noise, noise, distortion, and all other undesired transmitter
degradations with an exception for a select number of discrete spurs impacting a select number of subcarriers. MER
requirements are measured with a calibrated test instrument that synchronizes to the OFDMA signal, applies a
receive equalizer in the test instrument that removes MER contributions from nominal channel imperfections related
to the measurement equipment, and calculates the value. The equalizer in the test instrument is calculated, applied
and frozen for the CM testing. Receiver equalization of CM linear distortion is not provided; hence this is
considered to be a measurement of unequalized MER, even though the test equipment contains a fixed equalizer
setting.
7.4.12.3.3.1 Definitions
MER is defined as follows for OFDMA. The transmitted RF waveform at the F connector of the CM (after
appropriate down conversion) is filtered, converted to baseband, sampled, and processed using standard OFDMA
receiver methods, with the exception that receiver equalization is not provided. The processed values are used in the
following formula. No external noise (AWGN) is added to the signal.
The carrier frequency offset, carrier amplitude, carrier phase offset, and timing will be adjusted during each burst to
maximize MER as follows:
-
One carrier amplitude adjustment common for all subcarriers and OFDM symbols in burst.
One carrier frequency offset common for all subcarriers resulting in phase offset ramping across OFDM
symbols in bursts.
One timing adjustment resulting in phase ramp across subcarriers.
One carrier phase offset common to all subcarriers per OFDM symbol in addition to the phase ramp.
MERi is computed as an average of all the subcarriers in a mini-slot for the ith mini-slot in the OFDMA grant:
MERi (dB) = 10 log10
where:
2
1
1
=1 =1,
Eavg is the average constellation energy for equally likely symbols,

M is the number of symbols averaged,
N is the number of subcarriers in a mini-slot,
ej,k is the error vector from the jth subcarrier in the mini-slot and kth received symbol to the ideal
transmitted QAM symbol of the appropriate modulation order.
A sufficient number of OFDMA symbols shall be included in the time average so that the measurement uncertainty
from the number of symbols is less than other limitations of the test equipment.
MER with a 100% grant is defined as the condition when all OFDMA mini-slots and any legacy channels in the
transmit channel set are granted to the CM.
MER with a 5% grant is defined as the condition when less than or equal to 5% of the available OFDMA mini-slots
and no legacy channels have been granted to the CM.
7.4.12.3.3.2 Requirements
Unless otherwise stated, the CM MUST meet or exceed the following MER limits over the full transmit power
range, all modulation orders, all grant configurations and over the full upstream frequency range.
The following flat channel measurements with no tilt (Table 79) are made after the pre-equalizer coefficients have
been set to their optimum values. The receiver uses best effort synchronization to optimize the MER measurement.
10/29/13
CableLabs
73
CM-SP-PHYv3.1-I01-131029
Table 79 - Upstream MER Requirements (with Pre-Equalization)

Parameter
Value
MER (100% grant)
Each mini-slot MER 44 dB (Notes 1,2)
MER (5% grant)
Pre-equalizer constraints
Coefficients set to their optimum values
Table Notes:
1. Up to 5 subcarriers within the entire upstream bandwidth with discrete spurs may be excluded from the MER calculation if
they fall within transmitted mini-slots. These 5 spurs are the same spurs that may be excluded for spurious emissions and not
an additional or different set.
2. This value is to be met when Pload = Pload_min_set.
The following flat channel measurements (Table 710) are made with the pre-equalizer coefficients set to unity and
no tilt and the receiver implementing best effort synchronization. For this measurement, the receiver may also apply
partial equalization. The partial equalizer must not correct for the portion of the CMs time-domain impulse
response greater than 200 ns or frequency-domain amplitude response greater than +1 dB or less than -3dB from the
average amplitude.
Table 710 - Upstream MER Requirements (no Pre-Equalization)
Parameter
Value
MER (100% grant)
MER (5% grant)
Pre-equalizer constraints
Pre-equalization not used
Table Notes:
1. Up to 5 subcarriers within the entire upstream bandwidth with discrete spurs may be excluded from the MER calculation if they
fall within transmitted mini-slots. These 5 spurs are the same spurs that may be excluded for spurious emissions and not an
additional or different set.
2. This value is to be met when Pload = Pload_min_set.
7.4.13 Cable Modem Transmitter Output Requirements

The CM MUST output an RF Modulated signal with characteristics delineated in Table 711.
Table 711 - CM Transmitter Output Signal Characteristics
Parameter
Value
Frequency
Support and be configurable to:

5-42 MHz
5-65 MHz
5-85 MHz
5-117 MHz
5-204 MHz
NOT to cause harmful interference above these frequencies for any configured option
may support > 204 MHz
Signal Type
OFDMA
Maximum OFDMA Channel Bandwidth
96 MHz
Minimum Occupied Bandwidth
6.4 MHz for 25 kHz subcarrier spacing

10 MHz for 50 kHz subcarrier spacing
Number of Independently configurable

OFDMA channels
Minimum of 2
Subcarrier Channel Spacing
25 kHz, 50 kHz
FFT Size
50 kHz: 2048 (2K FFT); 1901 Maximum active subcarriers

25 kHz: 4096 (4K FFT); 3801 Maximum active subcarriers
Sampling Rate
102.4 MHz (96 MHz Block Size)
74
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Parameter
Value
FFT Time Duration
40 sec (25 kHz subcarriers)

20 sec (50 kHz subcarriers)
Modulation Type
BPSK, QPSK, 8-QAM, 16-QAM, 32-QAM, 64-QAM, 128-QAM, 256-QAM, 512-QAM,

1024-QAM, 2048-QAM, 4096-QAM
Bit Loading
Variable from mini slot to mini slot

Constant for subcarriers of the same type in the mini-slot
Support zero valued subcarriers (subcarrier nulling) per profile and minislot.
Pilot Tones
14 data patterns and 2 subslot patterns, mini-slot subcarrier size and length dependent
- see Section 7.4.16.
Cyclic Prefix Options
Samples
96
128
160
192
224
256
288
320
384
512
640
Windowing Size Options
Samples
sec
0
0
32
0.3125
64
0.625
96
0.9375
128
1.25
160
1.5625
192
1.875
224
2.1875
Raised cosine absorbed by CP
Level
MUST be capable of transmitting a total average output power of 65 dBmV

MAY be capable of transmitting a total average output power greater than 65 dBmV
Output Impedance
75 ohms
Output Return Loss
> 6 dB 5-fmax MHz (42/65/85/117/204 MHz)

> 6 dB fmax 1218 MHz
> 6 dB fmax 1.794 GHz for CMs that can receive up to 1.794 GHz
Connector
F connector per [ISO/IEC-61169-24] or [SCTE 02]
sec
0.9375
1.25
1.5625
1.875
2.1875
2.5
2.8125
3.125
3.75
5.0
6.25
7.4.14 CMTS Receiver Input Power Requirements

The CMTS Upstream Demodulator MUST operate with an average input signal level, including ingress and noise to
the upstream demodulator, up to 31 dBmV.
The CMTS MUST be settable according to Table 712 for intended received power normalized to 6.4 MHz of
bandwidth.
The CMTS Upstream demodulator MUST operate within its defined performance specifications with received bursts
within the ranges defined in Table 712 of the set power.
Table 712 - Upstream Channel Demodulator Input Power Characteristics
10/29/13
Modulation
Minimum Set Point
Maximum Set Point
Range
QPSK
8-QAM
16-QAM
32-QAM
64-QAM
-4 dBmV
-4 dBmV
-4 dBmV
-4 dBmV
-4 dBmV
10 dBmV
10 dBmV
10 dBmV
10 dBmV
10 dBmV
-9 / +3
-9 / +3
-9 / +3
-9 / +3
-9 / +3
CableLabs
75
CM-SP-PHYv3.1-I01-131029
Modulation
Minimum Set Point
Maximum Set Point
Range
128-QAM
256-QAM
512-QAM
1024-QAM
2048-QAM
4096-QAM
0 dBmV
0 dBmV
0 dBmV
0 dBmV
7 dBmV
10 dBmV
10 dBmV
10 dBmV
10 dBmV
10 dBmV
10 dBmV
10 dBmV
-9 / +3
-9 / +3
-3 / +3
-3 / +3
-3 / +3
-3 / +3
7.4.15 Ranging
Ranging in DOCSIS 3.1 is divided into three steps, as illustrated in Figure 716:
Initial ranging is used by the CMTS to identify a new admitting CM and for coarse power and timing ranging.
Fine ranging is used after initial ranging has been completed, to fine-tune timing and power.
Wideband probing is used during admission and steady state for pre-equalization configuration and periodic TX
power and time-shift ranging.
Initial
Ranging
Fine
Ranging
Fine
Ranging
Wideband
Probe
Admission
Steady State
Wideband
Probe
Fine
Ranging
Wideband
Probe
Wideband
Probe
Figure 716 - Ranging Steps
7.4.15.1 Initial Ranging

This section specifies the initial ranging scheme for DOCSIS 3.1.
7.4.15.1.1 Initial Ranging Zone
The initial ranging zone consists of N by M contiguous minislots in the upstream frame. N and M are configured by
the CMTS.
Minislots in the initial ranging zone do NOT carry pilots; all the FFT grid points in the initial ranging zone are used
for the initial ranging signal, as illustrated in Figure 717.
76
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Figure 717 - Initial Ranging Zone
7.4.15.1.2 Initial Ranging Signal

The initial ranging signal consists of a preamble sequence and a data part, as illustrated in Figure 718. The data part
is O-INIT-RNG-REQ as described in [DOCSIS MULPIv3.1].
Ngb/2
N*Q
subcarriers
Nir
Mir symbols
Preamble
Unique ID
Ngb/2
Tp Trigger point for the transmission of the initial ranging signal
Figure 718 - Initial Ranging Signal
The preamble sequence is a BPSK binary sequence configured by the CMTS and sent by the CM. The length of the
sequence is configured by the CMTS, and the bits contained in the sequence are configured by the CMTS.
The data portion of the initial ranging signal is the O-INIT-RNG-REQ message as described in [DOCSIS
MULPIv3.1]. It is composed of a 6-byte MAC address, plus a 1-byte downstream channel ID and 24 CRC bits. It is
LDPC (128,80) encoded and randomized as described in the sections below. The 24 CRC bits are generated as
described in Annex E.
The preamble sequence and the O-INIT-REG-REQ are duplicated and sent in a special structure of pair of symbols
with identical BPSK content as described in Figure 719.
10/29/13
CableLabs
77
CM-SP-PHYv3.1-I01-131029
Figure 719 - Initial Ranging Admission Slot Structure
A block diagram of the initial ranging signal processing in the transmitter is described in the following figure:
Preamble
sequence
Preamble
generator
Symbol
duplication
iFFT
UID
FEC encoder
CP/CS
addition
and
windowing
Randomizer
Figure 720 - Block Diagram of Initial Ranging Transmitter Processing
7.4.15.1.3 Preamble Construction

The CMTS MUST configure the BPSK Preamble sequence and its length Lp, with the limitations described in
Section 7.4.15.2.6 and the number of subcarriers, Nir, to be used for the transmission of the initial ranging signal.
The CMTS MUST allocate minislots for the initial ranging signal, comprising of the number of subcarriers Nir and
an appropriate guard band.
The CM MUST construct the preamble part of the initial ranging signal by converting the preamble sequence bits
into BPSK symbols. The preamble is comprised of Mir symbols each with Nir subcarriers.
The CM MUST convert the first Nir*Mir bits in the preamble sequence into Nir*Mir BPSK symbols in the following
order: The first Nir BPSK symbols are written to the Nir subcarriers of the first preamble symbol starting from the
lowest subcarrier, the next Nir BPSK symbols to the Nir subcarriers of the second preamble symbol and the last Nir
BPSK symbols to the Nir subcarriers of the last (the Mir) preamble symbol.
7.4.15.1.4 FEC for the Initial Ranging Data
The CM MUST encode the 80 bit O-INIT-RNG-REQ message using the LDPC (128,80) encoder as described
below.
A puncturing encoder consists of two steps. The first step encodes the input bit sequence with an encoder of the
mother code. The second step, called puncturing step, deletes one or more bits from the encoded codeword.
The mother code is a rate (160, 80) binary LDPC code. A parity check matrix of the mother code is represented
by Table 713, where sub-matrix size (lifting factor) L = 16, see Section 7.4.4.2 for the compact definition of parity
check matrix.
78
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Table 713 - (160,80) LDPC code Parity Check Matrix
1
2
0
6
12
11
1
9
8
13
10
14
3
11
12
15
2
10
-
7
14
3
0
9
14
-
12
11
-
7
10
-
4
5
Let the information bits sent to the mother code encoder be denoted by ( a0 ,, a79 ) and let the encoder output be
denoted by (a0 ,, a79 , b80 ,, b159 ) , where b80 ,, b159 are parity-check bits. The bits to be deleted by the
puncturing step are (also see Figure 721)
Period 1: 16 consecutive bits a0 ,, a15
Period 2: 16 consecutive bits b144 ,, b159
a0 a1 a79
80
Mother
Code
Encoder
a0 a79b80 b159
a16 a17 a79b80b81 b143
16
(1st period)
16
( 2nd period)
80 information bits
a16 a17 a78 a79 b80 b142b143
64
64
Puncture
Figure 721 - LDPC Two-Period Puncturing Encoder for Initial Ranging FEC
7.4.15.1.5 Padding and Randomizing

The CM MUST pad and randomize the 128 encoded bits as described below.
The CM MUST calculate the number of symbols required to transmit the INIT-RNG-REQ message as follows:
Nuid_sym = ceiling (128/Nir) where Nir is the number of subcarriers allocated for the INIT-RNG-REQ
message.
CM MUST pad the remaining bits with ones if the total number of bits (Nbits = Nuid_sym*Nir) is greater than 128.
The CM MUST randomize the 128 encoded bits and the padding bits as described in Section 7.4.5. The randomized
bits are converted to BPSK symbols as defined in (BPSK constellation) and are appended to the preamble sequence
for transmission.
The CM MUST add the BPSK symbols to the data part of the initial ranging signal in the following order: The first
Nir BPSK symbols written to the Nir subcarriers of the first symbol of the data part, the next Nir BPSK symbols to
the next data symbol, until all BPSK symbols are written vertically symbol by symbol. First BPSK symbol is written
to the lowest indexed subcarrier of a data symbol. Unfilled subcarriers in the last symbols are padded with 1s.
7.4.15.1.6 Symbol duplicating cyclic prefix and windowing
Each Initial Ranging OFDMA symbol is repeated twice. A cyclic prefix of Ncp samples is appended before the first
repeated OFDMA symbol. A cyclic suffix of Nrcp (Nrcp = Ncp + Nrp) samples is appended after the second repeated
OFDMA symbol.
10/29/13
CableLabs
79
CM-SP-PHYv3.1-I01-131029
Symbol X0
N
Repeat OFDM symbol
Symbol X0
Symbol X0
N
Add Prefix and Postfix
Cyclic
Prefix X0
Symbol X0
Ncp
Ncp
Nrcp =
Ncp + Nrp
Symbol X0
Cyclic
Postfix X0
Nrcp
Add Windowing
Symbol X0
Symbol X0
Symbol X+1
2*(N+Ncp)
Figure 722 - Initial Ranging Symbol Pair Structure
Table 714 - Cyclic Prefix and RollOff Samples for Initial Ranging
Roll-Off Samples (Nrp)
96
128
160
192
224
256
288
320
384
512
640
96
128
160
192
224
224
224
224
224
224
224
7.4.15.2 Fine Ranging

This section describes Fine ranging operations for the CM transmitter.
Fine ranging is used by the CMTS for the second step of the admission of a new CM process, following successful
initial ranging. During this step, a fine ranging signal is transmitted by a new CM joining the network, according to
transmission parameters provided by the CMTS. When it receives the fine ranging signal, the CMTS is able to finetune the joining CM's transmission power and transmission timing.
At the end of the fine ranging step, the CMTS can assign transmission opportunities to the new CM, using optimal
transmission power, without interfering with coexisting transmitters on the same OFDMA frame.
80
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
7.4.15.2.1 Fine ranging signal

Figure 723 illustrates a fine ranging signal.
Figure 723 - Fine Ranging Signal
Fine ranging is a narrowband signal integrated into a single data OFDMA frame. It is comprised of two parts: a
BPSK preamble sequence of one pair of preamble symbols (as defined in Section 7.4.15.1.2 for the initial ranging),
and 34 bytes of FEC-encoded data spread over two or more OFDMA symbols. The data part of the fine ranging
signal is QPSK-modulated and FEC encoded. The data part has a similar structure to the duplicated pair of symbols
(refer to Section 7.4.15.1.2 for the initial ranging data structure).
The CM MUST transmit the fine ranging signal when allocated to it, with the following configurable parameters:
Time shift
TX power
Number of minislots allocated to the fine ranging signal (number of minislots incorporate the fine ranging
signal plus the required guardband as described in Figure 724)
Number of subcarriers for the fine ranging signal
PRBS sequence.
The CM MUST use the first portion of the preamble sequence defined for the Initial Ranging signal for the BPSK
PRBS sequence of the fine ranging.
7.4.15.2.2 Transmission of The Fine Ranging Signal
The CM MUST duplicate the OFDMA symbols at the output of the IFFT as described in Section7.4.15.1.6, adding a
Cyclic Prefix to symbols 2n, and a Cyclic Suffix to symbols 2n+1, for n=0, 1, 2,
The CM MUST duplicate the OFDM symbols at the output of the IFFT, add cyclic prefix and suffix and apply
windowing as described in Section 7.4.15.1.6.
The CM MUST add the Cyclic Prefix as described in Section 7.4.11, using the same CP value used for all other
symbols.
The CM MUST add a Cyclic Suffix as described in Section 7.4.15.1.6, and the value of the Cyclic Suffix MUST be
equal to the Cyclic Prefix value.
The CM MUST use the Roll-off value specified in Section 7.4.11; the Roll-off value MUST be the same as that for
all other symbols except Initial Ranging Symbols.
Note:
The Roll-off value used for fine ranging may be different from the corresponding value used for Initial
Ranging.
The CM MUST start to transmit the fine ranging signal one symbol (including the cyclic prefix) after the start time
of the OFDMA frame.
10/29/13
CableLabs
81
CM-SP-PHYv3.1-I01-131029
Figure 724 - Fine Ranging Signal Transmission
The CM MUST transmit the fine ranging signal with guardband of Ngb/2 subcarriers from each side of the allocated
subcarriers. The CM calculates the number of subcarriers required for the guardband (Ngb) as follows:
Ngb = m*Q Nfr,
where
m: is the number of minislots allocated for the fine ranging
Nfr : is the number of subcarriers as configured by the CMTS.
The CM MUST transmit zero valued subcarriers in the guardband.
Figure 724 describes the fine ranging signal with M minislots of Q sub-carriers and K symbols and Ngb sub-carriers
for the guardband.
A block diagram of the initial ranging signal processing in the transmitter is described in Figure 725.
Figure 725 - Fine Ranging Transmitter Processing
7.4.15.2.3 Fine Ranging FEC

The CM MUST encode the 34 bytes of fine ranging information data using (362,272) shortening and puncturing
LDPC encoder.
Shortening and puncturing encoder consists of three steps. In this step, the shortening step, one or more information
bits are filled with 0 and the rest are filled with input bits. Then all information bits are encoded using the mother
code matrix. After mother code encoding, the zero filled bits are deleted. The puncturing step is as described in
Section 7.4.4.3.
The mother code is a rate 3/5 (480,288) binary LDPC code. A parity check matrix of the mother code is represented
by Table 715, where sub-matrix size (lifting factor) L = 48, see Section 7.5.4.1 for the compact definition of parity
check matrix.
82
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Table 715 - (480, 288) LDPC Code Parity Check Matrix
16
28
5
18
1
42
2
18
28
36
18
40
9
11
16
18
40
39
25
0
38
9
47
34
16
8
-
38
2
-
19
7
32
Denote the information bits sent to the mother code encoder by (a0 ,, a287 ) and let the encoder output being
(a0 ,, a287 , b288 ,, b479 ) , where b288 ,, b479 are parity-check bits. Then the shortening and puncturing steps
can be described as follows:
The shortening step fills 0 to 16 consecutive bits starting at position 272, i.e., let a 272 = a 273 = = a 287 = 0 . The
rest 272 bits i.e., a0 ,, a 271 , are fine ranging information data.
The bits to be deleted by the puncturing step are:
Period 1: 54 consecutive bits a0 , a1 ,, a53
Period 2: 48 consecutive bits b432 , b433 ,, b479
a0 a1 a271
Mother
code
Encoder
a0 a1 a271 00
000
16
272
272 information
a0 a271 000b288 b479
Shortenin
fil
a54 a55 a271b288b289 b431

54
(1st period)
a0 a271 b288 b479
192
a54 a55 a271 b288b289 b431
48
( 2nd period)
218
144
Puncturin
Figure 726 - Shortening and Puncturing Encoder for The Fine Ranging FEC
7.4.15.2.4 Padding, Randomizing and Interleaving

The CM MUST calculate the total number of data bits that can be transmitted in the fine ranging signal as follows:
Number_of_allocated_bits = Nfr* floor((K-4)/2)*2.
If the number of allocated bits is greater than 362, the CM MUST pad the 362 bits output from the LDPC encoder
with ones so that the encoded data and the pad bits equal the Number_of_allocated_bits.
The CM MUST randomize the data and padding bits as descried in Section 7.4.5.
The CM MUST interleave the randomized data as described in Section 7.4.6.
The CM MUST transmit zero valued subcarriers in all symbol times not used for the preamble, data and pad bits.
Note:
If K is an even number, the CM transmits K-2 symbols in the fine ranging signal (including the preamble), if
K is an odd number, the CM transmits K-3 symbols (including the preamble).
7.4.15.2.5 Exclusion Bandwidths

Transmission of the fine ranging signal around exclusion bands is allowed, using same processing as explained in
Section 7.4.15.2.2 with the same values of Nfr and Ngb.
Transmission with exclusion bands is illustrated in Figure 727.
10/29/13
CableLabs
83
CM-SP-PHYv3.1-I01-131029
Ngb/2
Fine ranging part 1
Nfr_1
Exclusion band
Nfr=Nfr_1+Nfr_2
Fine ranging part 2
Nfr_2
Ngb/2
Figure 727 - Fine Ranging and Exclusion Bands
7.4.15.2.6 Allowed Values and Ranges for Configuration Parameters

The CMTS MUST configure the fine ranging signal with the following limitations:
Maximum number of subcarriers for initial ranging MUST NOT exceed 64 subcarriers with 50 kHz subcarrier
spacing and 128 subcarriers with 25 kHz subcarrier spacing.
Maximum number of subcarriers for fine ranging MUST NOT exceed 256 subcarriers with 50 kHz subcarrier
spacing and 512 subcarriers with 25 kHz subcarrier spacing.
Maximum preamble sequence size for initial and fine ranging MUST NOT exceed 512 bits (64 Bytes) for 50
kHz and for the 25 kHz subcarrier spacing.
The maximum number of subcarriers for fine ranging including subcarriers in the exclusion zones MUST NOT
exceed 512 subcarriers.
Maximum number of preamble symbols (before duplication) for initial ranging MUST NOT exceed 8.
Maximum number of preamble symbols (before duplication) for fine ranging MUST be 1.
7.4.15.2.7 Power and Time Adjustments

Algorithms for power and time adjustments (such as number of fine ranging trials, frequency allocations, etc.) are
vendor-specific implementation.
7.4.15.3 Wideband Probing
Wideband probing is used during admission and steady state for pre-equalization configuration and periodic
transmission power and time-shift ranging.
7.4.15.3.1 Probing Frame
A probing frame consists of K contiguous probing symbols (OFDM symbols), where K is the number of symbols in
the minislot. The probing frame is aligned with the minislot boundaries in the time domain.
7.4.15.3.2 Probing Symbol Pilots
Probing symbol pilots are BPSK subcarriers, generated from the PRBS generation scheme described in
Section 7.4.15.3.3.
The CM MUST use the generation scheme detailed in Section 7.4.15.3.3 to generate 2048/4096 subcarriers for
2K/4K FFT.
The CM MUST use the same BPSK modulation for a specific subcarrier in all probing symbols.
The CM MUST transmit zero valued subcarriers in exclusion subcarriers.
84
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Probing symbol pilot i is always associated with the i-th subcarrier number, where:
i = 0, 1, ... , 2047 for 2K FFT
and
i = 0, 1, ... , 4095 for 4K FFT
(Subcarriers are numbered in ascending order starting from 0.)
7.4.15.3.3 PRBS generation scheme
The polynomial definition for the PRBS scheme is X12 + X9 + X8+ X5+ 1, where the seed is 3071. The period of the
PRBS is 212-1 bits, which is sufficient to create one probe symbol without repetitions. The sequence is illustrated in
Figure 728.
The CMs linear feedback shift register MUST be clocked after every subcarrier starting at subcarrier 0.
MSB
1
Delay
Delay
Delay
Delay
Delay
Delay
Delay
Delay
Delay
Delay
Delay
LSB
1
Delay
PRBS
SEQUENCE
Figure 728 - Polynomial Sequence for Pseudorandom Binary Sequence Generation
The PRBS sequence for 4K FFT is:

1 1 1 1 1 1 1 1 1 1 0 1 0 1 1 1 1 1 0 0 1 0 1 0 0 0 1 0 1 1
The PRBS sequence for 2K FFT is:
1 1 1 1 1 1 1 1 1 1 0 1 0 1 1 1 0 1 0 0 1 1 0 0 0 0 0 1 1 1
The PRBS sequence is mapped to the complex BPSK pilots as follows:
0 is mapped to a BPSK pilot of 1
1 is mapped to a BPSK pilot of -1
7.4.15.3.4 Probing Information
The CMTS MUST allocate a specific probing symbol within the probing frame and instruct the CM to transmit the
probing sequence in that symbol.
The CMTS MUST specify the probing symbol within the probing frame through the parameter "Symbol in Frame".
The CMTS MUST send three parameters to the CM: "st", "Start Subcarrier", and "Subcarrier Skipping".
The CM MUST support staggering pattern [DOCSIS MULPIv3.1] for probing , when the staggering bit "st" is set to
one, when "st" is set to zero, the staggering is off.
CMTS MUST define a probing pattern consisting of either the pilots from all the subcarriers of the assigned probing
symbol, or a set of pilots from scattered subcarriers of the assigned probing symbol. Please refer to section 6.4.4 in
[DOCSIS MULPIv3.1] for detailed probe mapping.
The range of "start subcarrier" is from 0 to 7. The range of "subcarrier skipping" is from 0 to 7. Figure 729 and
Figure 730 illustrate the use of these parameters.
10/29/13
CableLabs
85
CM-SP-PHYv3.1-I01-131029
The CM MUST use the start subcarrier and subcarrier skipping parameters to determine which subcarriers are to be
used for probing transmission, as follows:
The "start subcarrier" parameter is the starting subcarrier number.
The "subcarrier skipping" parameter is the number of subcarriers to be skipped between successive pilots.
"Subcarrier skipping" = 0 implies no skipping of subcarriers (i.e., all subcarriers in a single symbol belong to a
single transmitter).
A CM does not transmit the probing sequence in an excluded subcarrier. Excluded subcarriers are those subcarriers
in which no CM is allowed to transmit, generally because they are frequencies used by other systems (including
guard-band subcarriers).
subcarrier4095
subcarrier4094
subcarrier3
subcarrier2
subcarrier1
subcarrier0
Start subcarrier - 0
Subcarrier skipping - 0
Figure 729 - 4K FFT Example, All Subcarriers Used for Probing, no Skipping
subcarrier4095
subcarrier4094
subcarrier3
subcarrier2
subcarrier1
subcarrier0
Start subcarrier - 0
Subcarrier skipping - 1
Figure 730 - 4K FFT Example, Alternate Subcarriers Used for Probing
86
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
7.4.16 Upstream Pilot Structure

Pilots are used by the CMTS receiver to adapt to channel conditions and frequency offset.
DOCSIS 3.1 specifies two minislot types, differing in the number of subcarriers per minislot, 8- and 16-subcarrier
minislots.
Two types of minislots are defined for each minislot size: edge minislots and body minislots.
The CM MUST use an edge minislot as the first minislot in a transmission burst.
The CM MUST use body minislots for all other minislots in a transmission burst with the following two exceptions:
1.
The CM MUST use an edge minislot for the first minislot of an OFDMA frame that is not a zero valued
minislot.
2.
The CM MUST use an edge minislot for the first minislot after an exclusion band or after one or more
contiguous skipped subcarriers or after a zero valued minislot.
Figure 731 below describes the usage of edge and body minislots. Note that TX-2 is a one minislot burst
comprising of a single edge minislot.
Body
Body
Body
TX-3
Body
Edge
Edge
TX-2
Body
Body
TX-1
Body
Edge
Figure 731 - Edge and Body Minislots in a Transmission Burst
Pilots are subcarriers that do not carry data. Instead, a pilot subcarrier encodes a pre-defined BPSK symbol known to
the receiver (see Section 7.4.15.3.3). DOCSIS 3.1 also specifies complementary pilots. Complementary pilots are
subcarriers that carry data, but with a lower modulation order than other data subcarriers in the minislot. If the
modulation order used for data in the minislot is M, complementary pilots will use modulation order equal to the
maximum between M-4 and 1 (BPSK). For example if the bit loading in a minislot is 12, Complementary Pilots use
8 bits. If the bit loading is 4, Complementary Pilots will use BPSK. The CMTS receiver MAY use complementary
pilots to enhance its signal processing, such as to improve the accuracy of the center frequency offset acquisition.
For each minislot size, seven pilot patterns are defined.
Pilot patterns differ by the number of pilots in a minislot, and by their arrangement within the minislot. The different
pilot patterns enable the CMTS to optimize its performance (physical layer rate and pilot overhead) according to
different loop conditions and variations of SNR with frequency. Each pilot pattern defines edge and body minislots.
10/29/13
CableLabs
87
CM-SP-PHYv3.1-I01-131029
Two additional pilot patterns are specified for subslots (see Section 7.4.16.4 and Section 7.4.16.5); these are
required for both the CM and the CMTS.
The following sections describe the seven pilot patterns for each minislot size, and the pilot patterns for subslots.
7.4.16.1 Pilot Patterns for 8-Subcarrier Minislots
Figure 732 and Figure 733 define the pilot patterns for edge and body minislots with 8 subcarriers.
The CM MUST support pilot patterns 1-7.
The CMTS MUST support pilot patterns 1-4.
The CMTS SHOULD support pilot patterns 5-7.
In each figure, the horizontal axis represents OFDMA symbols, and the vertical axis represents the subcarriers. Each
square in a figure represents a subcarrier at a specific symbol time. Pilots are designated by "P" and complementary
pilots by "CP". All other subcarriers carry data with the modulation order of the minislot.
The figures show patterns for K between 6 and 16. For K>16 the complementary pilots are always located in the 14th
and 16th symbols, all symbols from the 17th symbol to the end of the frame carry data only. Pilot locations are the
same for any K.
Note:
This description assumes that the 8-subcarrier minislots are used with subcarrier spacing of 50 kHz. If the 8subcarrier minislots are used with subcarrier spacing of 25 kHz, the pilot pattern behavior with K is defined in
Section 7.4.16.2.
Pattern 1
Pattern 2
Pattern 3
Pattern 4
(K equal 6 to 16)
(K equal 6 to 16)
(K equal 6 to 16)
(K equal 6 to 16)
CP
CP
CP
CP
Body
minislots
CP
CP
CP
Edge
minislots
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
Figure 732 - Pilot Patterns 1-4 for Minislots with 8 Subcarriers
88
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Pattern 5
Pattern 6
Pattern 7
Pattern 7
(K equal 6 to 16)
(K equal 6 to 16)
(K equal 7 to 16)
(K equal 6)
CP
CP
CP
CP
P
Body
minislots
P
P
P
P
CP
CP
CP
CP
P
Edge
minislots
P
P
P
P
CP
CP
P
CP
CP
CP
7.4.16.2 Pilot patterns for 16-Subcarrier Minislots

Figure 734 and Figure 735 define the pilot patterns for minislots with 16 subcarriers
The CM MUST support pilot patterns 8-14.
The CMTS MUST support pilot patterns 8-11.
The CMTS SHOULD support pilot patterns 12-14.
The CMTS MUST configure minislots with 16 subcarriers to be used with 25 kHz subcarrier spacing with the
exception for RFoG.
The figures show patterns for K between 6 and 9. For K>9, the complementary pilots are always located in the 7th
and 9th symbols, all symbols from the 10th symbol to end of frame carry data only. Pilot locations are the same for
any K.
The horizontal axis in the figure represents OFDMA symbols, and the vertical axis represents subcarriers. Each
square in a figure represents a subcarrier at a specific symbol time. Pilots are designated by "P" and complementary
pilots by "CP". All other subcarriers carry data with the modulation order of the minislot.
10/29/13
CableLabs
89
CM-SP-PHYv3.1-I01-131029
Pattern 8
(K equal 6 to 9)
P
CP
Pattern 9
(K equal 6 to 9)
CP
CP
Pattern 10
(K equal 6 to 9)
P
CP
Body
minislots
CP
CP
CP
CP
P
Edge
minislots
CP
CP
CP
CP
CP
Pattern 11
(K equal 6 to 9)
CP
CP
CP
CP
CP
P
P
P
P
CP
CP
CP
CP
CP
CP
90
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Pattern 12
(K equal 6 to 9)
Pattern 13
(K equal 6 to 9)
P
CP
Pattern 14
(K equal 7 to 9)
CP
Pattern 14
(K equal 6)
P
CP
CP
Body
minislots
CP
CP
CP
CP
P
Edge
minislots
CP
CP
CP
CP
CP
7.4.16.3 Pilot Boosting

The CM MUST use higher power (pilot boost) when transmitting pilots and complementary pilots with pilot
patterns 5-7 and patterns 12-14, with the following exception:
The CM MUST use boosted power for the pilot and normal power for the complementary pilot when both are used
in the same symbol.
Pilots and complementary pilots are boosted by a factor of 3 in power (about 4.7 dB).
7.4.16.4 Pilot Patterns for 8-Subcarrier Subslots
Subslots are used to carry REQ messages which are always 7 bytes or 56 bits long. Data subcarriers are always
QPSK-modulated, and are not encoded by any FEC but are randomized using the randomizer described in Section
7.4.5.
Figure 736 depicts the pilot pattern for a subslot with 8 subcarriers.
The CM MUST support the pilot pattern for 8-subcarrier subslots.
The CMTS MUST support the pilot pattern for 8-subcarrier subslots.
Pilots are designated by "P", and no complementary pilots are used; all other subcarriers carry data with the
modulation order of the subslot.
10/29/13
CableLabs
91
CM-SP-PHYv3.1-I01-131029
P
P
P
Figure 736 - Pilot Pattern for Subslots with 8 Subcarriers
7.4.16.5 Pilot Patterns for 16-Subcarrier Subslots

Figure 737 depicts the pilot pattern for a subslot with 16 subcarriers.
The CM MUST support the pilot pattern for 16-subcarrier subslots.
The CMTS MUST support the pilot pattern for 16-subcarrier subslots.
Pilots are designated by "P", and no complementary pilots are used: all other subcarriers carry data with the
modulation order of the subslot.
Figure 737 - Pilot Pattern for Subslots with 16 Subcarriers
7.4.16.6 Content Transmitted by Pilots

Pilots use a PRBS sequence across frequency, over the whole range of 96 MHz. The PRBS sequence length is 1920
and 3840, for subcarrier spacing of 50 kHz and 25 kHz, respectively. Each subcarrier over the entire OFDMA
channel has its unique BPSK element. The series is generated as shown in Section 7.4.15.3.
The CM MUST use the first 2048 outputs of the generator to BPSK-modulate pilots with 50 kHz subcarrier spacing.
92
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
The CM MUST use the 4095 outputs of the generator to BPSK-modulate pilots with 25 kHz subcarrier spacing
starting from the first subcarrier.
7.4.17 Upstream Pre-Equalization
A CM MUST implement a linear pre-equalizer with a single complex coefficient per subcarrier.
The CMTS MUST direct a CM to pre-equalize its upstream transmission using CMTS-assigned pre-equalization
coefficients as a step in the ranging process.
The CMTS uses the CM's probe signal for pre-equalizer coefficient updates. The probes are described in Section
7.4.15.3. The message used to send information required for updating the pre-equalizer coefficients is described in
the Ranging Response (RNG-RSP) section of [DOCSIS MULPIv3.1].
The CMTS MAY specify the subcarriers (i.e., frequency range) over which coefficient updates is to be performed.
The CMTS MUST compensate for phase offset and phase slope in the coefficients if the probes are used for timing
offset measurements when calculating the pre-equalizer coefficient. Phase offset is compensated by averaging phase
over all subcarriers and subtracting it. The phase slope is compensated by subtracting the phase linear trend over all
subcarriers, after unwrapping 2*pi jumps.
The CMTS MUST have the ability to scale the transmission power per subcarrier when configuring the probe
transmission using the Range Response message.
The CM MUST scale its transmission power per channel when transmitting the probe as required by the CMTS in
the RNG-RSP message. The range of the scaling values is: 0 to [10log(skip+1)] dB. Skip is defined in Section
7.4.15.3.4.
The CM MUST use a default value of 1+j0 for all pre-equalizer coefficients. The RNG-RSP MAC message carries
the pre-equalization information, and may instruct the CM to either multiply the equalizer coefficients with the
existing coefficients (as described below), or to load them directly.
The CM MUST multiply the coefficients sent by the CMTS in the Range Response message with the existing
coefficients, to get the new coefficients, when instructed by the CMTS, as follows:
Ck(i+1)=Ck(i) * Ak(i)
where: Ck(i) is the pre-equalizer coefficient of the k-th subcarrier, as used in the last probe transmission, Ck(i+1) is
the updated pre-equalizer coefficient of the k-th subcarrier and Ak(i) is the update coefficient information received
from the CMTS as a response to the corresponding probe transmission. "*" indicates a complex multiplication.
The CM MUST normalize the new calculated coefficients as follows:
mean (abs (Ck)^2 ) = 1 (summation is over all k)
The CM MUST implement per subcarrier pre-equalization that will enable it to meet the performance requirements
specified in Section 7.4.12.
The CM MUST apply pre-equalization for all its data transmissions.
The CM MUST be able to transmit a probe signal with or without pre-equalization (all coefficients are 1+j*0) when
measuring upstream frequency response.
7.5
Downstream Transmit and Receive
7.5.1
Overview
This section specifies the downstream electrical and signal processing requirements for the transmission of OFDM
modulated RF signals from the CMTS to the CM.
This section is organized into subsections as summarized below:
10/29/13
CableLabs
93
CM-SP-PHYv3.1-I01-131029
7.5.1 Overview
Describes the structure of the DOCSIS 3.1 Physical Layer (PHY) specification sub-section on CMTS
transmitter requirements and CM receiver requirements.
7.5.2 Signal Processing
Informative description of the DOCSIS 3.1 downstream channel signal processing including forward error
correction, mapping of bits to constellation symbols, pilot insertion, Next Codeword Pointer insertion,
interleaving, exclusion subcarrier insertion, PHY Link Channel insertion, inverse discrete Fourier
transformation, pre-pending the cyclic prefix, and windowing.
7.5.3 Time and Frequency Synchronization
Defines CMTS transmitter requirements for providing time and frequency references for downstream OFDM
operation and CM requirements for recovering and synchronizing to time and frequency references.
7.5.4 Downstream Forward Error Correction
Defines CMTS transmitter and CM receiver requirements for forward error correction codeword creation,
shortening, encoding, decoding and interleaving.
7.5.5 Mapping Bits to QAM Constellations
Defines CMTS requirements for mapping serialized data bits from the FEC encoder to QAM constellations, bit
stream demultiplexing, cell word randomization, cell word mapping to constellations, bit loading, and inserting
Next Codeword Pointers, and defines CM and CMTS modulation format requirements.
7.5.6 Interleaving and De-interleaving
Defines CMTS requirements for time interleaving then frequency interleaving OFDM symbols and describes
impacts of interleaving on pilots, PHY link channel and subcarrier excluded spectral regions. Defines CM
requirements for frequency and time de-interleaving.
7.5.7 IDFT
Defines CMTS requirements for performing inverse discrete Fourier transform on the frequency-domain signal
to be transmitted to cable modems. Defines CM requirements for performing discrete Fourier transform on the
signal received from the CMTS.
Defines CMTS requirements for inserting cyclic prefixes and applying windowing to the output of the IDFT
and CM requirements for processing cyclic prefixes.
Defines CMTS transmitter requirements and limits for phase noise, modulation error ratio, spurious emissions,
frequency accuracy, and modulation timing jitter.
Defines CMTS transmitter requirements and limits for phase noise, modulation error ratio, spurious emissions,
distortion, and noise,
7.5.10 CMTS Transmitter Output Requirements
Defines CMTS transmitter electrical output requirements including output power level, impedance, channel
bandwidth and return loss.
7.5.11 Cable Modem Receiver Input Requirements
Defines CM receiver input requirements including channel bandwidth, number of FFT blocks, bit loading,
power level, impedance and return loss.
7.5.12
Cable Modem Receiver Capabilities
Defines CM receiver capabilities and error ratio performance requirements.
7.5.13 Physical Layer Link Channel (PLC)
Defines requirements for the structure, preamble modulation, location, content, and forward error correction for
the Physical Layer Link Channel.
94
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
7.5.14 Next Codeword Pointer

Defines requirements for bytes-to-bits mapping, forward error correction, FEC encoded bits to subcarriers
mapping, subcarrier placement, and randomization for the Next Codeword Pointer.
7.5.15 Downstream Pilot Patterns
Defines the structure of continuous and scattered pilots for downstream OFDM channels, describes the
signaling of the pilot structure in the PLC, and defines CMTS transmitter requirements for modulating and
boosting the power of pilots.
7.5.2
Signal Processing
Serial data signals received from the PHY-MAC Convergence Layer are received and processed by the PHY as
illustrated in Figure 738. This process yields OFDM symbols with 4096 subcarriers for the 4K FFT mode and 8192
subcarriers for the 8K FFT mode, with each symbol consisting of:
Data subcarriers
Scattered pilots
Continuous pilots
PLC subcarriers
Excluded subcarriers that are set to zero
This section briefly describes that process and provides links to the specific requirements for each process described
in this specification.
10/29/13
CableLabs
95
CM-SP-PHYv3.1-I01-131029
Forming
NCP
Constellation
Mapping
with
Randomization
Forward
Error
Correction
Convergence
Layer
NCP
Constellation Mapping
with Randomization
NCP
FEC
NCP Insertion
Bit-loading Profiles
& Scattered Pilot Map
Interleaving
Time
Interleaving
Continual Pilot
and Exclusion
Band Insertion
Frequency
Interleaving
IDFT
Cyclic Prefix
and
Windowing
Constellation
Mapping
with
Randomization
PLC FEC
Figure 738 - Downstream PHY Processing
96
CableLabs
10/29/13
7.5.2.1
CM-SP-PHYv3.1-I01-131029
Forward Error Correction (FEC) Encoding
The PHY begins processing incoming data by FEC encoding to data bits to form encoded codewords. Forward error
correction adds redundancy to the transmitted data; these redundant bits can be used by the receiver to detect and
correct errors in the transmission. For DOCSIS 3.1, FEC encoding applies a concatenated BCH-LDPC encoder,
based on [DVB-C2], and then shuffling the bits in a codeword via bit interleaving. Downstream forward error
correction is described in detail in Section 7.5.4.
7.5.2.2
Symbol Mapping to QAM Constellations
Once FEC encoded codewords have been created, the codewords are placed into OFDM symbols. Because each
subcarrier in an OFDM symbol can have a different QAM modulation, the codewords must first be demultiplexed
into parallel cell words; these cell words are then mapped into constellations based on the corresponding bit loading
pattern of the subcarrier's QAM constellation. In DOCSIS 3.1, QAM constellations extend from 16-QAM to 4096QAM (excluding 32-QAM), with both square and non-square constellations. This process is described in Section
7.5.4.1.1.
7.5.2.3
Scattered Pilot Placeholder Insertion
OFDM transmission requires the insertion of scattered pilots to enable channel estimation and equalization in the
receiver. While the insertion happens after time and frequency interleaving, since these pilots are not in the same
spectral location in every symbol, insertion of these scattered pilots disrupt the spectral location of the QAM data
subcarriers. To overcome this problem, place-holders for scattered pilot locations are inserted during the symbol
mapping process.
7.5.2.4
Next Codeword Pointer Insertion
Detecting where the next codeword begins in an OFDM symbol can be difficult: more than one codeword may map
into one OFDM symbol, the number of codewords per OFDM symbol may not be an integer, a codeword can
overflow from one OFDM symbol to another, and the codeword could be shortened. Therefore, the transmitter must
convey to the receiver all of the locations where a new codeword begins within an OFDM symbol. These Next
Codeword Pointers (NCPs) are encoded using another forward error correction method and are appended to OFDM
symbols. The process of encoding and inserting the NCP for DOCSIS 3.1 is discussed in Section 8.3.4.
7.5.2.5
Interleaving
These OFDM symbols, comprised of data subcarriers, scattered pilot placeholders, and NCPs, are then subjected to
time and frequency interleaving. Time interleaving mitigates the impact of burst noise, while frequency interleaving
mitigates the effect of ingress.
Time interleaving disperses the subcarriers of an input symbol over a set of output symbols, based on the depth of
interleaving. Therefore, if an OFDM symbol is corrupted by a noise burst, this burst is spread over the symbols
when it is de-interleaved, thereby reducing the error correction burden on the decoder. The time interleaving process
is described in Section 7.5.6.
Frequency interleaving occurs after time interleaving. Frequency interleaving disperses subcarriers of the symbol
along the frequency axis; therefore, OFDM subcarriers impacted by narrowband ingress are distributed between
several codewords, reducing the number of errors in each codeword. The frequency interleaving process is described
in Section 7.5.6.
7.5.2.6
Insertion of Pilots and Exclusion Sub-Bands
When interleaving is complete, placeholders for continuous pilots are inserted. These will be subject to modulation
later, together with the placeholders already inserted for scattered pilots. Continuous pilots are pilots that occur at
the same subcarrier location in every symbol. These are needed for receiver synchronization.
Exclusion bands and excluded subcarriers are inserted next. Nothing is transmitted at these subcarrier locations. The
contiguous block of subcarriers allocated to the PHY Link Channel is also treated as an exclusion band at this stage;
10/29/13
CableLabs
97
CM-SP-PHYv3.1-I01-131029
this is a placeholder for the PLC that is filled later. The regions outside the bandwidth of the OFDM signal may also
be treated as exclusion bands.
Finally, the placeholders for scattered and continuous pilots are replaced with a BPSK pseudo-random sequence.
This process is described in Section 7.5.15.
7.5.2.7
Encoding and Insertion of the PLC
The PLC is constructed within the convergence layer in parallel with the functions already discussed, relating to the
main data channel. The PLC occupies the same contiguous set of subcarriers in every OFDM symbol. The PLC data
is encoded for error correction, and then mapped into 16-QAM PLC data subcarriers. The PLC data is not subjected
to the same time or frequency interleaving as the data; however they are block interleaved. The PLC is then inserted
in place of its placeholder in each symbol. This process is described in Section 7.5.6.3.
7.5.2.8
IDFT Transformation and Cyclic Prefix Insertion
In this stage each OFDM symbol is transformed into the time domain using a 4096-point or 8192-point inverse
discrete Fourier transform (IDFT). This 4096 or 8192 sample sequence is referred to below as the IDFT output. This
process is described in Section 7.5.7.
7.5.2.9
A segment at the end of the IDFT output is prepended to the IDFT, and this is referred to as the Cyclic Prefix (CP)
of the OFDM symbol. There are five possible values for the length of the CP and the choice depends on the delay
spread of the channel a longer delay spread requires a longer cyclic prefix.
For windowing purposes another segment at the start of the IDFT output is appended to the end of the IDFT output
the roll-off postfix (RP). There are five possible values for the RP, and the choice depends on the bandwidth of the
channel and the number of exclusion bands within the channel. A larger RP provides sharper edges in the spectrum
of the OFDM signal; however, there is a time vs. frequency trade-off. Larger RP values reduce the efficiency of
transmission in the time domain, but because the spectral edges are sharper, more useful subcarriers appear in the
frequency domain. There is an optimum value for the RP that maximizes capacity for a given bandwidth and/or
exclusion band scenario.
These topics are discussed in detail in Section 7.5.8.
7.5.3
Time and Frequency Synchronization
This section specifies the timing and frequency synchronization requirements for DOCSIS 3.1 CMTS transmitters
and CM receivers.
The purpose of this section is to ensure that the CMTS transmitter can provide proper timing and frequency
references for DOCSIS 3.1 downstream OFDM operation and that the CM receiver can acquire the system timing
and subcarrier from the downstream for proper DOCSIS 3.1 operation.
The CMTS downstream OFDM symbol and subcarrier frequency and timing relationship is defined in Section 7.3.3.
7.5.3.1
Downstream sampling rate
The CMTS MUST lock the 204.8 MHz Downstream OFDM Clock to the 10.24 MHz CMTS Master Clock (see
Table 71).
7.5.3.2
OFDM RF Transmission Synchronization
The CMTS MUST lock the Downstream OFDM RF transmissions to the 10.24 MHz CMTS Master Clock (see
Table 71).
98
CableLabs
10/29/13
7.5.3.3
CM-SP-PHYv3.1-I01-131029
Downstream OFDM Symbol Clock Jitter
The CMTS MUST adhere to the following double sideband phase noise requirements for the downstream OFDM
symbol clock over the specified frequency ranges:
< [-21 + 20*log (fDS /204.8)] dBc (i.e., < 0.07 nSec RMS) 10 Hz to 100 Hz
< [-21 + 20*log (fDS /204.8)] dBc (i.e., < 0.07 nSec RMS) 100 Hz to 1 kHz
< [-21 + 20*log (fDS /204.8)] dBc (i.e., < 0.07 nSec RMS) 1 kHz to 10 kHz
< [-4 + 20*log (fDS /204.8)] dBc (i.e., < 0.5 nSec RMS) 10 kHz to 100 kHz
< [2 + 20*log (fDS /204.8)] dBc (i.e., < 1 nSec RMS) 100 kHz to (fDS /2),
where fDS is the frequency of the measured clock in MHz.
The CMTS MUST use a value of fDS that is an integral multiple or divisor of the downstream symbol clock. For
example, an fDS = 409.6 MHz clock may be measured if there is no explicit 204.8 MHz clock available.
7.5.3.4
Downstream Timing Acquisition Accuracy
The downstream clock timing is defined with respect to downstream OFDM frame.
The CM MUST be able to adjust its clock to synchronize its own clock timing with OFDM downstream frame for
proper operation.
The CM MUST be able to acquire downstream clock timing from downstream traffic (pilots, preambles, or mixed
pilots, preambles, and data).
The CM MUST have a timing acquisition accuracy better than 1 sample (4.8828125 ns).
7.5.3.5
Downstream Carrier Frequency Acquisition Accuracy
The CM MUST be able to acquire the carrier frequency from downstream (pilots, preambles, or mixed pilots,
preambles and data).
7.5.3.6
Downstream Acquisition Time
The CM MUST achieve downstream signal acquisition (frequency and time lock) in less than 60s for a device with
no previous network frequency plan knowledge.
Nonetheless, it is expected that the CM would be able to achieve downstream acquisition in less than 30s.
7.5.4
Downstream Forward Error Correction
This section describes the downstream forward error correction scheme used for DOCSIS 3.1. It is based on [DVBC2] section 6.1, FEC Encoding; it is used here with the following modifications:
A codeword will be the size of the short FECFrame (16,200 bits); the "normal" FEC Frame (64,800 bits) is not
used.
Only the code rate 8/9 is used.
Support for non-square constellations (128-QAM, 512-QAM, and 2048-QAM) is introduced.
Support for mixed modulation codewords is introduced.
Support for codeword shortening is introduced.
Bit Interleaving for non-square constellations (128-QAM, 512-QAM, and 2048-QAM), mixed modulation
mode constellations and for shortened codewords is introduced.
Demultiplexing for non-square constellations (128-QAM, 512-QAM, and 2048-QAM), mixed modulation
mode constellations, and for shortened codewords is introduced.
Support for QPSK modulation is not required.
10/29/13
CableLabs
99
CM-SP-PHYv3.1-I01-131029
These changes are described in the following sections.

7.5.4.1
7.5.4.1.1
Definitions
Mixed-Modulation Codewords
Before downstream FEC can be defined, it is important to understand what a mixed-modulation codeword is, as
these codewords are handled differently. A mixed-modulation codeword belongs to a profile that does not use the
same modulation constellation for all sub-carriers of the OFDM symbol. Note that subcarrier zero bit loading is not
taken into account when determining if a codeword is a mixed-modulation codeword. In other words, if a profile has
the same modulation constellation (i.e., same bit loading profile) for all non-zero bit-loaded subcarriers of the
OFDM symbol, then the codewords of that profile are not considered to be mixed modulation.
As an example consider a profile in which even numbered subcarriers, excluding zero bit-loaded subcarriers (i.e.,
non-zero bit-loaded subcarriers 0, 2, 4, 6, etc.), are modulated with modulation A, and odd numbered subcarriers
(i.e., non-zero bit-loaded subcarriers 1, 3, 5, 7, etc.) are modulated with modulation B. This provides a bits/s/Hz
value that is the mean of the bits/s/Hz values of modulations A and B. Any codeword that belongs to that profile is a
mixed-modulation codeword.
In this example, if these subcarriers are modulated as shown in the following table, the modulation combinations
provide approximately 1.5 dB SNR granularity of additional spectral efficiency:
Table 716 - Mixed Modulation with 1.5 dB SNR Granularity
Modulation A
Modulation B
128-QAM
256-QAM
512-QAM
1024-QAM
2048-QAM
256-QAM
512-QAM
1024-QAM
2048-QAM
4096-QAM
Another example of the use of mixed-modulation codewords can be applied to the case of an OFDM channel at the
high frequency end with a significant tilt in SNR. In this case, a modulation profile for this part of the spectrum
could use four different QAM constellations covering the OFDM symbol: 1024-QAM, 512-QAM, 256-QAM and
64-QAM. All of the codewords to which this profile is applied would be of the mixed-modulation type.
It is important to note that a codeword is treated as being a mixed-modulation type even if all of the subcarriers have
the same modulation order; being of the mixed-modulation type is determined by the profile. For example, consider
a codeword of the above profile in which all the subcarriers happen to be 256-QAM. Despite the fact that all
subcarriers of this codeword have the same modulation, this codeword is treated as the mixed-modulation type since
it belongs to a mixed-modulation profile. It is necessary to do this because the FEC encoder has no knowledge as to
which subcarriers the codeword is going to be mapped while the encoding is being performed. Therefore, FEC
encoder operations are determined by the profile applied to the codeword only.
As a final example, consider a profile that consists of 75% 1024-QAM subcarriers and 25% zero-bit-loaded
subcarriers. In this case the codewords of that profile are not of the mixed-modulation type, since zero-bit-loaded
subcarriers are ignored when determining mixed-modulation type.
7.5.4.1.2
Codeword vs. FECFrame
[DVB-C2] uses the term FECFrame to refer to the bits of one LDPC encoding operation. In this specification, the
term codeword is used for the same concept.
7.5.4.2
FEC Encoding
[DVB-C2] section 6.1, FEC Encoding, describes the FEC encoding requirements for the CMTS transmitter. The
CMTS MUST meet the portion of [DVB-C2] section 6.1, FEC Encoding, as described below:
100
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
The CMTS MUST support the 8/9 code rate for the short codeword (Nldpc = 16,200 bits) only. Support for other code
rates and codeword sizes is not required.
The CMTS MUST support the FEC coding parameters specified in Table 717. This table is based on Table 3(b),
from [DVB-C2].
Table 717 - Coding Parameters (for Short Codewords Nldpc = 16,200 and Code Rate 8/9)
LDPC Code
Rate
BCH Uncoded
Block Size Kbch
BCH Coded
Block Nbch
LDPC Uncoded
Block Size Kldpc
LDPC Coded Block

Size Nldpc
8/9
14,232
14,400
14,400
16,200
7.5.4.2.1
Outer Encoding (BCH)
[DVB-C2] section 6.1.1, Outer Encoding (BCH), details the outer encoding requirements for normal and short
codewords (FECFrames). For the CMTS, only short codewords are required. The CMTS MUST meet the outer
encoding requirements for short FECFrames specified in [DVB-C2] section 6.1.1, Outer Encoding (BCH).
7.5.4.2.2
Inner Coding (LDPC)
[DVB-C2] sections 6.1.2, Inner Encoding, and 6.1.2.2, Inner Coding for Short FECFrame, detail the inner coding
requirements for short codewords. For DOCSIS 3.1 codewords, the CMTS MUST meet the inner coding
requirements for short codewords and code rate 8/9 specified in [DVB-C2] sections 6.1.2, Inner Encoding, and
6.1.2.2, Inner Coding for Short FECFrame.
7.5.4.2.3
Support for Codeword Shortening
Codeword shortening is used for two purposes:
Create shortened codewords when there is insufficient data to fill complete codewords.
Achieve strong burst noise protection
The full FEC block size for the FEC code rate of 8/9 is provided in Table 717.
Codeword shortening is accomplished by shortening the uncoded block size in the BCH Uncoded Block Size column
of Table 717. Note that the number of parity bits remains the same; there is no shortening of the parity bits either in
the BCH or in the LDPC.
When a shortened codeword is needed, the CMTS MUST complete the codeword shortening process described here.
There are six overall steps to the codeword shortening process:
1.
Prepending zero bits (BCH) to the data
2.
BCH encoding
3.
Removing the prepended zero bits
4.
Appending zero bits (LDPC) to the data
5.
LDPC Encoding
6.
Removing the appended zero bits
This is done in both the BCH encoder and the LDPC encoder, as shown in Figure 739.
10/29/13
CableLabs
101
CM-SP-PHYv3.1-I01-131029
K bits
Append
Kldpc-K-168
zero bits
Prepend
Kbch-K
zero bits
Kldpc
Kbch
LDPC
Encoder
Nbch
BCH
Encoder
Nldpc
Remove
Kldpc-K-168
zero bits
Remove
Kldpc-K-168
zero bits
K+168
K+168+Nldpc-Kldpc
Figure 739 - Codeword Shortening Process
The zero bit padding process is shown in more detail in Figure 740.
102
CableLabs
10/29/13
Payload bits
CM-SP-PHYv3.1-I01-131029
Padding for
BCH Encoding
Kbch-K
zeros
BCH Encoded
Kbch-K
zeros
168
BCH Padding
Removal
168
Padding for
LDPC Encoding
168
(Kldpc-K-168) zero bits
After LDPC
Encoding
168
(Kldpc-K-168) zero bits
LDPC Padding
Removal
168
1800
Parity bits
1800
Parity bits
Length of codeword = (K + 168 + 1800) bits

Figure 740 - Padding Process
7.5.4.2.3.1
Codeword Shortening for Strong Burst Noise Protection
Although the primary purpose of codeword shortening is to support scenarios in which there is insufficient data to
fill complete codewords, codeword shortening can also be used to provide signal protection in strong burst noise
conditions. A lower code rate such as 7/9 has better burst noise capabilities than the 8/9 code rate. Through
codeword shortening, it is possible to achieve the equivalent of a 7/9 code rate.
For example, the 16200-bit block can be shortened by 8096 bits. The number of parity bits remains unchanged at
1800. Hence, this shortened codeword will have a block size of 8104 with 6304 information bits and 1800 parity
bits; this produces an effective code rate of approximately 7/9 (6304/8104 = 0.777887463). When the receiver
receives this shortened codeword, it will pad the shortened 8096 bits with zeros to create a 16200-bit rate 8/9
codeword and decode it using the rate 8/9 decoder.
7.5.4.3
7.5.4.3.1
Bit Interleaving
Bit Interleaving for Non-Shortened Codewords
For non-shortened codewords that are not of the mixed-modulation type the CMTS MUST apply parity interleaving,
followed by column-twist interleaving as detailed in [DVB-C2] section 6.1.3, Bit Interleaver, with the number of
rows, columns and column twisting parameters specified in this section.
10/29/13
CableLabs
103
CM-SP-PHYv3.1-I01-131029
The number of rows and columns of the Bit Interleaver are specified by Table 718.
Table 718 - Bit Interleaver Structure
Modulation
Rows Nr
Columns Nc
16-QAM
64-QAM
128-QAM
256-QAM
512-QAM
1024-QAM
2048-QAM
4096-QAM
8192-QAM
16384-QAM
2025
1350
2315
2025
1800
810
1473
675
1247
1158
8
12
7
8
9
20
11
24
13
14
Since 16,200 is not divisible by 7, 11, 13 and 14, for 128-QAM, 2048-QAM, 8192-QAM and 16384-QAM
constellations, the CMTS MUST append zeros after parity interleaving and prior to column-twist interleaving at the
end of the block: 5 zero bits for 128-QAM, 3 zero bits for 2048-QAM, 11 zero bits for 8192-QAM and 12 zero bits
are added after the 16,200th bit. Thus, an extended block of 16,205 bits, 16,203 bits, 16,211 bits and 16,212 bits will
be interleaved by the column-twist interleaver for 128-QAM, 2048-QAM, 8192-QAM and 16384-QAM,
respectively.
For non-shortened codewords that are not of the mixed-modulation type, the CMTS MUST serially write the data
bits into the column-twist interleaver column-wise, and serially read out row-wise, where the write start position of
each column is twisted by tc, as specified in Table 719 and Table 720.
Table 719 - Column Twisting Parameter tc (columns 0 - 11)
Twisting Parameter tc
Codeword
Modulation
Type
Columns
Nc
Col. 0
10
11
16-QAM
64-QAM
128-QAM
256-QAM
512-QAM
1024-QAM
2048-QAM
4096-QAM
8192-QAM
16384-QAM
8
12
7
8
9
20
11
24
13
14
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
0
0
0
0
0
0
0
2
0
2
0
0
0
0
2
1
2
2
1
3
2
0
0
0
2
7
2
2
7
5
2
0
0
0
2
20
2
3
20
6
2
3
0
0
2
20
3
3
20
7
2
3
0
2
2
21
3
21
9
2
4
1
3
4
3
11
5
4
1
5
4
6
5
4
1
8
6
7
5
4
2
8
6
7
5
2
9
8
104
Codeword
Modulation
Type
Columns
Nc
12
13
14
15
16
17
18
19
20
21
22
23
16-QAM
64-QAM
128-QAM
256-QAM
512-QAM
1024-QAM
2048-QAM
4096-QAM
8192-QAM
16384-QAM
8
12
7
8
9
20
11
24
13
14
5
2
9
9
7
3
9
7
7
-
7
9
-
7
9
-
8
9
-
8
10
-
10
10
-
10
-
10
-
10
-
11
-
CableLabs
10/29/13
7.5.4.3.2
CM-SP-PHYv3.1-I01-131029
Bit Interleaving for Non-Shortened Mixed Modulation Codewords
To support non-shortened mixed-modulation codewords, the bit interleaver specified in [DVB-C2] has been
modified. For non-shortened mixed-modulation codewords, the CMTS MUST apply parity interleaving followed by
column-twist interleaving as detailed in [DVB-C2] section 6.1.3, Bit Interleaver, and in the following discussion.
Because specific columns of the bit de-interleaver cannot be mapped to specific bits of the QAM constellation,
column twisting interleaving is used over all 24 columns.
For non-shortened mixed-modulation codewords, the CMTS MUST serially write the data bits into the column-twist
interleaver column-wise, and serially read out row-wise, where the write start position of each column is twisted by
tc, as specified in Table 721 and Table 722.
Codeword
Modulation
Type
MixedModulation
Columns
Nc
Col.
0
24
10
11

Codeword
Modulation
Type
MixedModulation
7.5.4.3.3
Columns
Nc
Col.
12
24
13
14
15
16
17
18
10
19
10
20
10
21
10
22
10
23
11
Bit Interleaving for Shortened Codewords
Shortened codewords fall into one of the following three types:

1.
Square modulation
2.
Non-square modulation
3.
Mixed modulation
The CMTS MUST interleave all types of shortened codewords as described in this section.
The CMTS MUST interleave the 1800 parity bits as described in [DVB-C2] section 6.1.3, Bit Interleaver.
Because the shortened codeword can be quite small, it is possible that the entire codeword could map to one column
of the interleaver and hence not get interleaved. To avoid this, the CMTS MUST use the maximum number of
columns in the bit interleaver (24) on shortened codewords. The number of rows that would be occupied by the
shortened codeword is given by the following equation:
_ =
+ 1968
24
If (K+1968) is not divisible by 24, then the last column will only be partially filled by the encoded shortened
codeword, as illustrated in Figure 741. For shortened codewords, the CMTS MUST fill the unfilled part of the last
column with bits that are labeled as "unused". For shortened codewords, the CMTS MUST discard these "unused"
bits in the memory read operation described below.
10/29/13
CableLabs
105
CM-SP-PHYv3.1-I01-131029
24 columns
11
Row_Count
Unused
Figure 741 - Bit De-interleaver Block for a Shortened Codeword
For shortened codewords, the CMTS MUST serially write the data bits into the column-twist interleaver columnwise, and serially read out row-wise, where the write start position of each column is twisted by tc, as specified in
Table 723 and Table 724. For shortened codewords, the CMTS must fill any unfilled bits of the last column with
bits marked "unused". The CMTS must write the first bit of the last column beginning from the 12th location since
tc=11 for column 24. If there are any bits left over after writing in the last location of the column, these bits must be
written beginning from the top of the column. For shortened codewords, the CMTS must discard any bits labeled as
"unused" during the process of reading along the rows of the two-dimensional array.
Codeword
Type
Shortened
Columns
Nc
Col.
0
24
10
11

Codeword
Type
Shortened
7.5.4.4
Columns
Nc
Col.
12
24
13
14
15
16
17
18
10
19
10
20
10
21
10
22
10
23
11
Downstream Receiver FEC Processing
Downstream data is encoded for FEC by the CMTS. The CM MUST decode the FEC-applied codeword to correct
for any bit errors introduced by noise and interference in the transmission medium. This process is discussed in this
section.
The FEC decoder at the CM operates on the QAM subcarriers of OFDM symbols to generate an error corrected bit
stream. In addition, the decoder generates error statistics such as codeword error ratios.
The FEC decoding process is shown in Figure 742.
106
CableLabs
10/29/13
LLR
De-Mapper
CM-SP-PHYv3.1-I01-131029
LLR
De-Interleaver
LDPC
Decoder
BCH
Decoder
Figure 742 - FEC Decoding Process
The receiver FEC decoder consists of the following components:
The log-likelihood ratio (LLR) de-mapper processes one OFDM subcarrier at a time from the OFDM symbol
and generates the LLRs for all bits of the QAM constellation, as defined by the bit-loading profile for the
specific subcarrier. For example, if the subcarrier is 1024-QAM, the LLR de-mapper will generate 10 LLRs for
the subcarrier and the values of the LLRs are implementation specific. The LLR de-interleaver operates on the
LLRs. This is the inverse of the bit-interleaver that has been applied by the CMTS transmitter, described in
Section 7.5.2.5. Note that the receiver operates on LLRs and not on bits.
The LDPC decoder decodes the 16200-bit LDPC codeword (or a shortened codeword). LDPC decoding is
implemented using an iterative algorithm that uses message passing between the bit nodes and the check nodes
of the Tanner graph of the LDPC code. If the CMTS has transmitted a shortened codeword (e.g., when the
payload is not large enough to fill a complete codeword), the receiver augments the shortened codeword to full
size with LLRs corresponding to zero-valued bits. The receiver then decodes the codeword using the 16200-bit
LDPC decoder, discarding the augmented bits.
The BCH decoder generates an error corrected bit stream. The BCH decoder is also required to operate on
shortened codewords.
An error monitor determines codeword error ratios for reporting and troubleshooting.
7.5.5
Mapping Bits to QAM Constellations
This section describes the method used in DOCSIS 3.1 to map symbols onto QAM constellations. It is based on
[DVB-C2] section 6.2, Mapping Bits onto Constellations, and is used here with the following modifications:
Parameters for mapping bits onto non-square constellations have been added
Parameters for mapping bits of shortened codewords onto all constellation types have been added
Parameters for mapping bits of mixed-modulation codewords onto all constellation types have been added
As described in [DVB-C2] section 6.2, Mapping Bits onto Constellations, the CMTS MUST map each codeword to
a sequence of QAM constellation values by:
Demultiplexing the input bits into parallel cell words
Mapping these cell words into constellation values
The mapping of bits to QAM constellation is carried out using the three sequential operations depicted in
Figure 743.
10/29/13
CableLabs
107
CM-SP-PHYv3.1-I01-131029
De-multiplexing
Bit-stream to
Cell Words
Mapping Cell
Words to QAM
Constellation
Points
Randomization
Figure 743 - Bits to QAM Constellation Mapping
The CMTS MUST use the number of bits per cell MOD, as defined in Table 725, when bit mapping codewords to
constellations.
For non-shortened codewords that are not of the mixed-modulation type, the CMTS MUST use the Number of
Output Data Cells defined in Table 725 when bit mapping codewords to constellations.
Table 725 - Parameters for Bit-Mapping onto Constellations
Modulation
Mode
MOD
Number of Output
Data Cells
16384-QAM
8192-QAM
4096-QAM
2048-QAM
1024-QAM
512-QAM
256-QAM
128-QAM
64-QAM
16-QAM
14
13
12
11
10
9
8
7
6
4
1158
1247
1350
1473
1620
1800
2025
2315
2700
4050
For the cases of mixed-modulation codewords and shortened codewords, the number of output symbols per LDPC
block remains an integer. For both mixed-modulation codewords and shortened codewords, the CMTS MUST pad
the end of the LDPC block with zero bits to produce an integer number of bits in the final QAM symbol. The CM
MUST discard zero pad bits in the received symbol. This is described in further detail in the following sections.
7.5.5.1
Modulation Formats
The CMTS modulator MUST support 16-QAM, 64-QAM, 128-QAM, 256-QAM, 512-QAM, 1024-QAM, 2048QAM, and 4096-QAM.
The CMTS modulator MAY support 8192-QAM and 16384-QAM.
The CM demodulator MUST support 16-QAM, 64-QAM, 128-QAM, 256-QAM, 512-QAM, 1024-QAM, 2048QAM, and 4096-QAM.
The CM demodulator MAY support 8192-QAM and 16384-QAM.
7.5.5.2
7.5.5.2.1
Bit-to-Cell Word Demultiplexer

Non-shortened Codewords
For non-shortened codewords that are not of the mixed-modulation type, the CMTS MUST demultiplex the bit
stream stream vi from the bit interleaver into Nsubstreams sub-streams, using the value of Nsubstreams as defined in Table
726 and the description following that table.
108
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Table 726 - Number of Sub-Streams in Demultiplexer
Modulation
Mode
Number of SubStreams, Nsubstreams
16-QAM
64-QAM
128-QAM
256-QAM
512-QAM
1024-QAM
2048-QAM
4096-QAM
8192-QAM
16384-QAM
8
12
7
8
9
20
11
24
13
14
Bit-to-cell word demultiplexing is illustrated in Figure 744.
b0,0 b0,1 b0,2

b1,0 b1,1 b1,2
v0 v1 v2
Bit to Cell Word

Demultiplexer
bNsubstreams-1,0 bNsubstreams-1,1 bNsubstreams-1,2
Figure 744 - Bit-to-Cell Word Demultiplexer
For 16-QAM, 64-QAM, 256-QAM ,1024-QAM and 4096-QAM bit-to-cell word demultiplexing has to be carried
out as described in [DVB-C2] section 6.2.1, Bit to Cell Word Demultiplexer.
Bit-to-cell word demultiplexing is defined as a mapping of the bit-interleaved input bits, vdi, onto the output bits
, , where:
vdi is the input to the demultiplexer;
di is the input bit number;
e is the demultiplexer sub-stream index (0 e < Nsubstreams), which depends on (di modulo Nsubstreams), as
defined in Table 727 through Table 731;
is the output cell number from the demultiplexer;
, is the output from the demultiplexer.
Table 727 - Parameters for Demultiplexing of Bits to Sub-Streams for

8/9 Code Rate with 128-QAM
10/29/13
Input bit number,

di mod Nsubstreams
Output bit number, e
CableLabs
109
CM-SP-PHYv3.1-I01-131029

Input bit number,

di mod Nsubstreams

Input bit number,

di mod Nsubstreams
10
Output bit
number, e
10

Input bit number,

di mod Nsubstreams
10
11
12
12
11
10

Input bit number,

di mod Nsubstreams
10
11
12
13
Output bit
number, e
13
12
10
11
For example, in the case of 128-QAM there will be 7 substreams at the output of the bit-to-cell word demultiplexer.
The first 7 bits at the input to the demultiplexer are sent to sub-streams 6, 5, 4, 1, 2, 3 and 0, in that order. The next 7
input bits are also mapped in that order. The cell words are defined from the demultiplexer output as:
0, . . 1, = [0, . . 1, ]
Note that the non-shortened LDPC codeword size is not divisible by 7. However, with reference to the section on bit
interleaving, it is seen that for 128-QAM the size of the non-shortened codeword has been extended to become a
multiple of 7 through zero-padding. The same comments are applicable to non-shortened 2048-QAM, 8192-QAM
and 16384-QAM codewords.
7.5.5.2.2
Shortened Codewords and Mixed-Modulation Codewords
It must be emphasized that shortened codewords can have square modulation, non-square modulation or may be of
the mixed-modulation type. The CMTS MUST bypass the bit-to-cell demultiplexer and apply the bit-to-cell
mapping described in this section for all types of shortened codewords, as well as for non-shortened mixed
modulation codewords.
When the bit-to-cell word demultiplexer is bypassed, the bit-to-cell mapping becomes:
Cell 0:
0,0 . . 01,0 = 0 01
Cell 1: 0,1 . . 11,1 = 0 0 +11

etc
110
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
The modulation assigned to cells 0 and 1 in the previous equations correspond to the mod values given by Table 7
25. The first cell has the modulation corresponding to mod0 and the second cell has the modulation corresponding to
mod1. This modulation is defined by the bit loading pattern assigned to the profile to which this codeword belongs.
This mapping is simply a case of partitioning the interleaved bit stream to blocks of bits of size mod0, mod1, ,
modLAST, where the sequence {mod0, mod1, , modLAST} is given by the bit loading pattern of the profile to which
this codeword belongs.
Let modLAST correspond to the bit loading of the last cell of the sequence. It is possible that the shortened and/or
mixed-modulation codeword at the output of the bit interleaver might not have sufficient bits to complete this cell.
In this case zero-padding of the input bit stream has to be used for cell completion.
7.5.5.3
Randomization
The CMTS MUST randomize cell words of data subcarriers, NCP subcarriers and PLC subcarriers, just before
mapping these onto QAM constellations, as described in this section.
The CMTS MUST also introduce BPSK-modulated subcarriers for the following subcarriers during the
randomization process, as described in this section.
a)
Zero-bit-loaded subcarriers of the codewords of individual profiles
b) Zero-bit-loaded subcarriers in the NCP segment

c)
Zero-bit-loaded subcarriers that are introduced to complete the symbol
NCP and zero bit-loading are described in Section 7.5.5.5.

The wordlength ( ) of a cell word ranges from 4 bits for 16-QAM to 14 bits for 16384-QAM.
For 16-QAM to 4096-QAM the CMTS MUST randomize each cell word through a bit-wise exclusive-OR operation
with the least significant bits (LSBs) of the 12-bit register D0 of the linear feedback shift register (LFSR)
shown in Figure 745.
0 .. 1 = 0 . . 1 (0 [0]. . 0 [ 1])
For 8192-QAM the CMTS MUST randomize the 13 bits of the cell word through a bit-wise exclusive-OR operation
with the 12 bits of register D0 and the LSB of register D1 of Figure 745, as given below:
(0 .. 12 ) = (0 . . 12 ) (0 [0]. . 0 [ 1] 1 [0])
For 16384-QAM the CMTS MUST randomize the 14 bits of the cell word through a bit-wise exclusive-OR
operation with the 12 bits of register D0 and the 2 LSBs of register D1 of Figure 745, as given below:
(0 .. 13 ) = (0 . . 13 ) (0 [0]. . 0 [ 1] 1 [0] 1 [1])
NCP subcarrier cell words are 2-bit for QPSK, 4-bit for 16-QAM or 6-bit for 64-QAM. The CMTS MUST
randomize these through bit-wise exclusive-OR operation with the 2, 4 or 6 LSBs of the 12-bit register D0.
The CMTS MUST set the zero-bit-loaded subcarriers in the data segment and NCP segment to the BPSK
modulation given by LSB of register D0.
0 = 0 [0]
The CMTS MUST clock the LFSR once, after each of the previous operations.
10/29/13
CableLabs
111
CM-SP-PHYv3.1-I01-131029
D1
12-bit
D0
11
12-bit
Figure 745 - Linear Feedback Shift Register for Randomization Sequence
The LFSR is defined by the following polynomial in [212 ].

2 + + 11
The [212 ] is defined through polynomial algebra modulo the polynomial:

12 + 6 + 4 + + 1
Each 12-bit [212 ] element is a polynomial of with a maximum degree of 11. The coefficient of 0 is referred to
as the LSB and the coefficient of 11 is referred to as the MSB.
This LFSR is initialized to the hexadecimal numbers given below:

D0 = "555"
D1 = "AAA"
This initialization is carried out at the beginning of an OFDM symbol, synchronized to the preamble of the PLC.
Since the PLC subcarriers are inserted after time and frequency interleaving and data subcarriers are randomized
before time and frequency interleaving, the following explanation is provided about how randomization is
synchronized to the PLC.
Note that the first subcarrier of an OFDM symbol passes through the time interleaver arm with zero delay. Therefore
the LFSR is initialized when this subcarrier is part of the OFDM symbol following the last OFDM symbol carrying
the PLC preamble. Hence LFSR is initialized once for every 128 OFDM symbols.
The first subcarrier referred to previously can be a data subcarrier or a scattered pilot placeholder because both of
these are time interleaved. If it is a data subcarrier then the cell word of that data subcarrier is randomized with the
initialized values of D0 and D1, namely hexadecimal "555" and "AAA". After that the LFSR is clocked once. If the
first subcarrier mentioned previously is a scattered pilot placeholder the LFSR is initialized but it is not clocked.
This is because the LFSR is clocked only after each data or NCP subcarrier (including zero-bit-loaded subcarriers).
7.5.5.4
Cell Word Mapping into I/Q Constellations
The CMTS MUST modulate each randomized cell word (0 . . 1 ) from the randomizer described in Section
7.5.5.3 using a BPSK, QPSK, 16-QAM, 64-QAM, 128-QAM, 256-QAM, 512-QAM, 1024-QAM, 2048-QAM,
4096-QAM, 8192-QAM or 16384-QAM constellation as described in Section 7.4.7.
7.5.5.5
Transmitter Bit Loading for Symbol Mapping
All subcarriers of an OFDM symbol may not have the same constellation; the constellation for each subcarrier is
given in a table that details the bit loading pattern. This bit-loading pattern may change from profile to profile. This
section describes how the bits to symbol mapping is performed, with reference to a bit-loading pattern, in the
presence of interleaving, continuous pilots, scattered pilots and excluded subcarriers.
112
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Excluded subcarriers are subcarriers that are forced to zero-valued modulation at the transmitter. Subcarriers are
excluded to prevent interference to other transmissions that occupy the same spectrum as the DOCSIS 3.1 OFDM
transmission, for example, to accommodate legacy channels. Subcarriers are also excluded outside of the active
OFDM bandwidth.
Excluded subcarriers are common to all profiles. The non-excluded subcarriers are referred to as active subcarriers.
Active subcarriers are never zero-valued. The notation () is used here to define the set of excluded subcarriers.
This set will never be empty because there are always excluded subcarriers at the edges of the OFDM channel.
Continuous pilots are pilots that occur at the same frequency location in every OFDM symbol. The notation () is
used here to define the set of continuous pilots.
The PLC resides in a contiguous set of subcarriers in the OFDM channel. The CMTS adds the PLC to the OFDM
channel after time and frequency interleaving; the CM extracts the PLC subcarriers before frequency and time deinterleaving. These subcarriers occupy the same spectral locations in every symbol. The notation () is used here to
define the set of PLC subcarriers.
For bit loading, continuous pilots and the PLC are treated in the same manner as excluded subcarriers; hence, the set
of subcarriers that includes the PLC, continuous pilots and excluded subcarriers is defined as:
() = () () ()
The subcarriers in the set () do not carry data (PLC carry signaling information). The other subcarriers that do
not carry data are the scattered pilots. However, scattered pilots are not included in the set () because they do
not occupy the same spectral locations in every OFDM symbol.
The modulation order of the data subcarriers is defined using bit-loading profiles. These profiles include the option
for zero bit-loading. Such subcarriers are referred to as zero-bit-loaded subcarriers and are BPSK modulated using
the randomizer LSB, as described in Section 7.5.5.3.
All active subcarriers with the exception of pilots are transmitted with the same average power. Pilots are
transmitted boosted by a factor of 2 in amplitude (approximately 6 dB).
Scattered pilots do not occur at the same frequency in every symbol; in some cases scattered pilots will overlap with
continuous pilots. If a scattered pilot overlaps with a continuous pilot, then that pilot is no longer considered to be a
scattered pilot. It is treated as a continuous pilot.
Because the locations of scattered pilots change from one OFDM symbol to another, the number of overlapping
continuous and scattered pilots changes from symbol to symbol. Since overlapping pilots are treated as continuous
pilots, the number of scattered pilots changes from symbol to symbol.
The following notation is used here:
: The total number of subcarriers in the OFDM symbol, equaling either 4096 or 8192
: The number of continuous pilots in an OFDM symbol
: The number of scattered pilots in an OFDM symbol
: The number of excluded subcarriers in an OFDM symbol
: The number of PLC subcarriers in an OFDM symbol

: The number of data subcarriers in an OFDM symbol
The values of , , and do not change from symbol to symbol for a given OFDM template; the values of
and change from symbol to symbol.
The following equation holds for all symbols:
= + + + +
The value of N is 4096 for 50 kHz subcarrier spacing and 8192 for 25 kHz subcarrier spacing. From this equation it
is clear that ( + ) is a constant for a given OFDM template. Therefore, although the number of data subcarriers
( ) and the number of scattered pilots ( ) in an OFDM symbol changes from symbol to symbol, the sum of these
10/29/13
CableLabs
113
CM-SP-PHYv3.1-I01-131029
two numbers is invariant over all symbols. Interleaving and de-interleaving are applied to the set of data subcarriers
and scattered pilots of size = + .
7.5.5.5.1
Bit Loading
The bit loading pattern defines the QAM constellations assigned to each of the 4096 or 8192 subcarriers of the
OFDM transmission. This bit loading pattern can change from profile to profile. Continuous pilot locations, PLC
locations and exclusion bands are defined separately, and override the values defined in the bit-loading profile. Let
the bit loading pattern for profile be defined as (), where:
k is the subcarrier index that goes from 0 to (N-1)
N is either 4096 or 8192
() {0, 4, 6, 7, 8, 9, 10, 11, 12, 13, 14}. A value of 0 indicates that the subcarrier k is zero-bit-loaded. Other
values indicate that the modulation of subcarrier k is QAM with order 2 () .
Let the sequence { (), = 0, 1, , ( 1), } be arranged as NI consecutive values of another

sequence:
(), = 0, 1, ,
( 1)
Given the locations of the excluded subcarriers, continuous pilots and the PLC in the OFDM template, it is possible
to obtain the bit-loading pattern () that is applicable only to spectral locations excluding excluded subcarriers,
continuous pilots, and PLC subcarriers. However, note that () does contain the spectral locations occupied by
scattered pilots; these locations change from symbol to symbol.
It is more convenient to define bit loading profiles in the domain in which subcarriers are transmitted. It is in this
domain that signal-to-noise-ratios of subcarriers are calculated. Furthermore, defining the bit-loading patterns in the
transmission domain allows significant data compression to be achieved, because a relatively large number of
contiguous spectral locations can share the same QAM constellation.
Although the bit loading pattern is defined in the domain in which subcarriers are transmitted, the bit loading is not
applied in that domain. Bit loading is applied prior to interleaving, as shown in Figure 746. Hence there is a
permutation mapping of subcarriers, defined by the interleaving function, between the domain in which bit loading
is applied to subcarriers and the domain in which subcarriers are transmitted.
Bits to QAM
Subcarrier &
OFDM
Symbol
Mapping
Receiver
Time &
Frequency
De-Interleaving
Transmitter
Time &
Frequency
Interleaving
Bit-Loading
Profiles
OFDM
Symbol
QAM
Subcarrier
De-Mapping
into LLRs
Bit-Loading
Profiles
Figure 746 - Bit Loading, Symbol Mapping, and Interleaving
The excluded subcarriers, PLC subcarriers, and continuous pilots are excluded from the processes of interleaving
and de-interleaving; scattered pilots and data subcarriers are subject to interleaving and de-interleaving. Hence, the
total number of subcarriers that pass through the interleaver and de-interleaver is = ( + ) and this number
does not change from symbol to symbol.
114
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
The interleaver introduces a 1-1 permutation mapping on the subcarriers. Although interleaving consists of a
cascade of two components, namely time and frequency interleaving, it is only frequency interleaving that defines
the mapping . This is because time interleaving does not disturb the frequency locations of subcarriers.
The corresponding permutation mapping applied at the receiver de-interleaver is 1 .
In order to perform bit-loading, it is necessary to work out the bit loading pattern at the node at which it is applied,
i.e., at the input to the interleavers. This is given by:
() = 1 ( ())
Since the time interleaver does not change the frequency locations of subcarriers, the sequence () is obtained by
sending { (), = 1, 2, , 1} through the frequency de-interleaver.
Note that () gives the bit-loading pattern for subcarriers. Yet, some of these subcarriers are scattered pilots
that have to be avoided in the bit-loading process. Hence, a two-dimensional binary pattern (, ) is used to
identify subcarriers to be avoided during the process of bit-loading. Because the scattered pilot pattern has a
periodicity of 128 in the time dimension, this binary pattern also has periodicity 128 in the column dimension j.
(, ) is defined for = 0, 1, , ( 1) and for = 0, 1, , 127
The process to create the binary pattern (, ) begins with the transmitted scattered pilot pattern defined in Section
7.5.6. There are two scattered pilot patterns, one for 4K FFTs and the other for 8K FFTs; both patterns are defined in
reference to the preamble of the PLC and have a periodicity of 128 symbols.
The CMTS executes the following steps to obtain the pattern (, ):
1.
Define a two-dimensional binary array (, ) in the subcarrier transmitted domain that contains a one for
each scattered pilot location and zero otherwise:
(, ), for = 0, 1, , 1 and for = 0, 1, , 127
Here, the value of N is either 4096 or 8192. The first column of this binary sequence corresponds to the
first OFDM symbol following the preamble of the PLC.
2.
3.
4.
Exclude the rows corresponding to excluded subcarriers, continuous pilots, and PLC from the twodimensional array (, ) to give an array (, ). The number of rows of the resulting array is and the
number of columns is 128.
Pass this two-dimensional binary array (, ) through the frequency de-interleaver and then the time deinterleaver, with each column treated as an OFDM symbol. After the 128 columns of the pattern have been
input into the interleaver, re-insert the first M columns, where M is the depth of the time interleaver. This is
equivalent to periodically extending (, ) along the dimension and passing (128+M) columns of this
extended sequence through the frequency de-interleaver and the time de-interleaver.
Discard the first M symbols coming out of the time de-interleaver and collect the remaining 128 columns
into an array to give the binary two-dimensional array (, ) of size ( 128).
For bit loading the CMTS accesses the appropriate column of the binary pattern bit (, ) together with the
appropriate bit loading profile (). If the value of the bit (, ) is 1, the CMTS MUST skip this subcarrier and
move to the next subcarrier. This subcarrier is included as a placeholder for a scattered pilot that will be inserted in
this subcarrier location after interleaving. After each symbol the column index has to be incremented modulo 128.
The CMTS MUST use this binary two-dimensional array (, ) of size ( 128) in order to do bit-loading of
OFDM subcarriers, as described earlier in this section.
The corresponding operation in the CM is de-mapping the QAM subcarriers to get Log-Likelihood-Ratios (LLRs)
corresponding to the transmitted bits. This operation, described below, is much simpler than the mapping operation
in the transmitter.
The scattered pilots and data subcarriers of every received symbol are subjected to frequency and time deinterleaving. The scattered pilots have to be tagged so that these can be discarded at the output of the time and
frequency de-interleavers. This gives subcarriers for every OFDM symbol. The CM accesses theses deinterleaved subcarriers together with the bit-loading pattern () to implement the de-mapping of the QAM
10/29/13
CableLabs
115
CM-SP-PHYv3.1-I01-131029
subcarriers into LLRs. If the subcarrier happens to be a scattered pilot, then this subcarrier, as well as the
corresponding value (), is skipped and the CM moves to the next subcarrier ( + 1).
7.5.5.5.2
NCP Insertion
Next Codeword Pointers (NCPs) point to the beginning of codewords in a symbol, counting only data subcarriers.
The format of an NCP is described in Section 8.3.4, Next Codeword Pointer, which also describes the FEC applied
to the NCP. Each FEC encoded NCP is 48 bits wide. NCPs may be modulated using QPSK, 16-QAM or 64-QAM
and this modulation is signaled by PLC. In addition to the NCPs carrying next codeword pointers, there will also be
a NCP carrying the CRC for all the NCPs of the symbol. The CRC is generated as described in Annex E. For 8K
FFT node, there will be NCPs in each symbol. For the 4K FFT, there will be one set of NCPs per pair of symbols.
As the NCPs are constructed while the OFDM symbols are being constructed, the NCPs are inserted in the opposite
direction to data and beginning from the opposite end. Data is inserted beginning from the low frequency towards
the high frequency end. The NCPs are inserted from the high frequency end towards the low frequency end.
Note that subcarriers in each symbol are subjected to the data and NCP mapping operation. These subcarriers
consist of data subcarriers and scattered pilot place-holder subcarriers as described in the preceding section. During
the course of mapping data or NCP subcarriers, if a scattered pilot placeholder is encountered, this is skipped.
The figure given below shows an OFDM symbol comprising a Data segment, an NCP segment and a "Filler"
segment. "Filler" subcarriers have to be inserted into the OFDM symbol when there is no data to be transmitted, or
when the number of codewords has exceeded the upper limit or when it is not possible to begin a new codeword
because of insufficient space to include a NCP. These filler subcarriers are zero-bit-loaded.
The CMTS MUST use zero-bit-loaded filler subcarriers when there is no data to be transmitted or when the number
of codewords has exceeded the upper limit or when it is not possible to begin a new codeword because of
insufficient space to include a NCP.
Data segment contains codewords belonging to several profiles. Some of the subcarriers may be zero-bit-loaded in
some of the profiles. The NCP also has a profile. This profile allows some of the subcarriers in the NCP segment to
be zero-bit-loaded. Note that the NCP modulation is a constant for given OFDM transmission. It does not change
from subcarrier to subcarrier.
Note that throughout the symbol there can be scattered pilot placeholders. These have to be skipped during the
insertion of data subcarriers, NCP subcarriers or filler subcarriers. Moreover, these have to be tagged before sending
the NI subcarriers through the time and frequency interleavers. Scattered pilots will be inserted in their place with the
appropriate BPSK modulation before the data is transmitted.
116
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
NCP Subcarriers including zero-bitloaded subcarriers and scattered pilot

placeholders
Filler subcarriers that are
zero-bit-loaded and scattered
pilot placeholders
Total of
NI subcarriers
Data subcarriers including zero-bitloaded subcarriers and scattered

pilot placeholders belonging to one
or more profiles
Frequency
Figure 747 - NCP Insertion
7.5.6
Interleaving and De-interleaving
To minimize the impacts of burst noise and ingress on the DOCSIS signals, time and frequency interleaving are
applied to OFDM symbols in the following order: time interleaving, then frequency interleaving. These interleaving
methods are discussed in this section.
The time interleaver is a convolutional interleaver that operates in the time dimension on individual subcarriers of a
sequence of OFDM symbols. The time interleaver does not change the frequency location of any OFDM subcarrier.
A burst event can reduce the SNR of all the subcarriers of one or two consecutive OFDM symbols; the purpose of
the time interleaver is to disperse these burst-affected OFDM subcarriers between M successive OFDM symbols,
where M is the interleaver depth. This dispersion distributes the burst-affected subcarriers uniformly over a number
of LDPC codewords.
The frequency interleaver works along the frequency dimension. The frequency interleaver changes the frequency
locations of individual OFDM subcarriers; latency is not introduced, except for the data store and read latency. The
aim of frequency interleaving is to disperse ingress, e.g., LTE that affects a number of consecutive subcarriers over
the entire OFDM symbol. Frequency interleaving distributes the burst-affected subcarriers over a number of LDPC
codewords.
The CMTS first applies a time interleaver to an OFDM symbol worth of NI subcarriers to get a new set of NI
subcarriers. These NI subcarriers are made up of ND data subcarriers and NS scattered pilots.
= +
It must be noted that although ND and NS are not the same for every OFDM symbol, the value of NI is a constant for
all OFDM symbols in a given system configuration. The value of NI is a function of the channel bandwidth, number
of excluded subcarriers, number of PLC subcarriers and the number of continuous pilots. The CMTS then subjects
these NI subcarriers to frequency interleaving. The value of NI does not exceed 7537 for 8K FFT mode and 3745 for
the 4K FFT mode.
Note that both time and frequency interleaving are applied only to data subcarriers and scattered pilots. Continuous
pilot, subcarriers that have been excluded (used to support legacy channels in spectral regions, for example) and the
subcarriers of the physical layer link channel (PLC) are not interleaved. The CMTS MUST NOT interleave
continuous pilots, excluded subcarriers or the subcarriers of the PLC.
10/29/13
CableLabs
117
CM-SP-PHYv3.1-I01-131029
7.5.6.1
Time Interleaving
The CMTS MUST time interleave as described in this section. The CMTS MUST time interleave after OFDM
symbols have been mapped to QAM constellations and before they are frequency interleaved.
The time interleaver is a convolutional interleaver that operates at the OFDM subcarrier level. If the depth of the
interleaver is M, then there are M branches, as shown in Figure 748.
Figure 748 - Time Interleaver Structure
The CMTS MUST support a maximum value of M equal to 32 for 20 s symbol duration (50 kHz subcarrier
spacing) and 16 for 40 s symbol duration (25 kHz subcarrier spacing).
The CMTS MUST support all values of M from 1 to the maximum value of M (inclusive of both limits).
Each branch is a delay line; the input and output will always be connected to the same delay line. This delay line
will be clocked to insert a new subcarrier into the delay line and to extract a subcarrier from the delay line. Next, the
commutator switches at the input, and the output will move to the next delay line in the direction shown by the
arrows in Figure 738. After the delay line with the largest delay, the switch will move to the delay line with zero
delay.
The lowest frequency subcarrier of an OFDM symbol always goes through the branch with zero delay. Then the
commutator switch at input and the corresponding commutator switch at output are rotated by one position for every
new subcarrier.
The value of J is given by the following equation:
= ( )
Here, NI is the number of data subcarriers and scattered pilots in an OFDM symbol. See Section 7.5.6.3 for details
on interleaving scattered pilots.
If NI were not divisible by M, all of the branches would not be filled. Therefore, "dummy subcarriers" are added to
the symbol to make the number of subcarriers equal to a multiple of M. The number of dummy subcarriers is given
by:

The dummy subcarriers are added for definition purposes only; at the output of the interleaver these dummy
subcarriers are discarded. An implementation will use a single linear address space for all the delay lines in Figure
738. Writing and reading dummy subcarriers will not be needed.
118
CableLabs
10/29/13
7.5.6.2
CM-SP-PHYv3.1-I01-131029
Frequency Interleaving
The CMTS MUST frequency interleave OFDM symbols as described in this section. The CMTS MUST frequency
interleave after OFDM symbols have been time interleaved.
The frequency interleaver works on individual OFDM symbols. Each symbol to be interleaved consists of NI
subcarriers. These NI subcarriers are made up of ND data subcarriers and NS scattered pilot placeholders. Although
ND and NS are not the same for every symbol, the value of NI is a constant for all OFDM symbols in a given system
configuration. See Section 7.5.6.3 for details on interleaving scattered pilots.
There is a 2-D store comprising 128 rows and K columns. If the number of data subcarriers and scattered pilots in
the OFDM symbol is NI, then the number of columns, K, is given by the following equation:
= (
)
128
If NI is not an integer multiple of 128, then the last column will only be partially filled during the frequency
interleaving process. The number of data subcarriers in the last column, C, is given by:
= 128( 1)
The frequency interleaver follows the following process; note that rows are numbered 0 to 127, and columns are
numbered from 0 to (K-1):
1.
Write the subcarriers along rows of the 2-D store. Rows are accessed in bit-reversed order. For example,
after writing in row 0, the next writing operation will be in row 64. This will be followed by writing in row
32 and so on. If the row number is less than C, then K subcarriers will be written in the row. Otherwise only
(K-1) subcarriers will be written. (If the number NI is an integer multiple of 128 then C will be zero. Then K
subcarriers will be written in every row.)
2.
Rotate columns 0 to (K-2) by an amount given by a 6-bit shift linear feedback (maximal length) shift
register. This shift register is initialized to a value of 17 at the start of each OFDM symbol. The final
column, which may be partially full, is not rotated.
3.
Read the columns in bit-reverse order, starting at column 0, then column bit-reverse(1), then column bitreverse(2), , ending at column bit-reverse(K-1). When K is not a power-of-2, bit-reverse(x), for x = 0,
, K-1, is defined by:
bit-reverse(x) = reverse_bits(x), if reverse_bits(x) < K; OR
x,
if reverse_bits(x) >= K
where reverse_bits(x) is the number obtained by reversing the order of the bits in the m-bit representation
of x, with m being the number of bits in K.
The structure of the two-dimensional store is shown in Figure 749.

0
K-2
K-1
127
Figure 749 - Two-Dimensional Block Structure
10/29/13
CableLabs
119
CM-SP-PHYv3.1-I01-131029
The linear feedback shift register is defined using the following equation in Galois field GF[26].
() = ( 1), = 1, , 127, (0) = 5 +
GF[26] is defined using the polynomial (6+ +1). As this is primitive, powers of will generate all 63 non-zero
elements of the field. This operation can be represented as the linear feedback shift register, depicted in Figure 750.
x5
x4
x3
x2
x1
x0
Figure 750 - Linear Feedback Shift Register
The binary number x[5:0] is used to rotate the columns. This number is initialized to 17 at the beginning of each
OFDM symbol. The column number 0 is rotated by 17; subsequent columns are rotated by values obtained by
clocking the shift register shown in Figure 750. The rotation applied to the first column is defined in Figure 751.
Subsequent rows are also rotated along the same direction.
111
112
113
Rotation by 17
125
108
126
109
127
110
Figure 751 - Frequency Interleaver Rotation Definition
Note that column (K-1) is not rotated, regardless of whether it is full: because all other columns are rotated by a nonzero amount, there is no need to rotate column (K-1). The C code for interleaver implementation is given in
Appendix I.
7.5.6.3
Interleaving Impact on Continuous Pilots, Scattered Pilots, PLC and Excluded Spectral
Regions
DOCSIS 3.1 transmissions contain continuous pilots for receiver synchronization and scattered pilots for channel
estimation. In addition, there could be nulled regions to accommodate legacy channels. There will also be a physical
layer link channel (PLC).
The CMTS interleaves scattered pilots and data subcarriers, but does not interleave continuous pilots, the PLC, and
subcarriers belonging to nulled regions. With respect to scattered pilots, it must be noted here that CMTS actually
interleaves the subcarriers that are tagged to act as placeholders for scattered pilots, since at the time of interleaving
120
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
the scattered pilots have not yet been inserted. The actual BPSK modulation to these placeholder subcarriers is
applied after interleaving as described in Section 7.5.15.
The CMTS inserts scattered pilot placeholders prior to time and frequency interleaving such that when these
placeholders get time and frequency interleaved, the resulting placeholders conform to the required scattered pilot
pattern described in Section 7.5.15.
To accomplish this, the CMTS has to retain a reference pattern for inserting scattered pilot placeholders prior to
interleaving. Since the scattered pilot pattern repeats every 128 symbols, this pattern is a (NI x 128) two-dimensional
bit pattern. A value of one in this bit-pattern indicates the location of a scattered pilot. The CMTS inserts data
subcarriers where this reference pattern has a zero and scattered pilot placeholders where this pattern has a one.
This reference pattern may be derived from the following procedure:
1.
In the time-frequency plane, create a two-dimensional bit-pattern of zeros and ones from the transmitted
"diagonal" scattered pilot patterns described in Section 7.5.15. This pattern has a periodicity of 128
symbols and has a value of one for a scattered pilot location and zero otherwise. Let the time axis be
horizontal and the frequency axis vertical.
2.
Delete all horizontal lines containing continuous pilots, excluded subcarriers, and PLC from the above
mentioned two-dimensional bit pattern; note the some scattered pilots could coincide with continuous
pilots. These locations are treated as continuous pilot locations.
3.
Send the resulting bit-pattern through the frequency de-interleaver and the time de-interleaver in
succession. This will give another two-dimensional bit pattern that has a periodicity of 128 symbols. The
appropriate 128-symbol segment of this bit-pattern is chosen as the reference bit pattern referred to above.
Note that the CMTS has to synchronize the scattered pilot pattern to the PLC preamble, as described in Section
7.5.15. This uniquely defines the 128-symbol segment that has to be used as the reference pattern.
Scattered pilots are not in the same subcarrier location in every symbol; hence some scattered pilots can coincide
with continuous pilots in some OFDM symbols. The size of the overlap between the set of scattered pilots and the
set of continuous pilots will change from symbol to symbol. As a result, the number of data subcarriers in a symbol
will not be the same for all OFDM symbols. Note that in the nomenclature used below, when a scattered pilot
coincides with a continuous pilot, then that pilot is referred to as a continuous pilot.
Although the number of data subcarriers can change from symbol to symbol, the number of data subcarriers and
scattered pilots are the same for every symbol. This is referred to as NI in this section. Let ND denote the number of
data subcarriers in a symbol and NS denote the number of scattered pilots in a symbol. These two parameters, i.e., ND
and NS, will change from symbol to symbol. However, the sum of these two, i.e., NI is a constant for a given system
configuration.
= +
Hence the number of OFDM subcarriers that are interleaved does not change from symbol to symbol. This is
important, because if not for this, the output of the convolutional time interleaver may have dummy or unused
subcarriers in the middle of interleaved OFDM symbols.
The insertion of continuous pilots, PLC and excluded regions happens after both time and frequency interleaving.
Interleaving data and scattered pilots together has another important advantage. This is to do with bit loading. A
transmitted profile is said to have non-uniform bit loading if the QAM constellation that is applied to subcarriers is
not constant over the entire frequency band. If the data subcarriers are interleaved and scattered pilots are added
later, then the data subcarriers will have to be shifted to accommodate the scattered pilots. This shift will be different
from symbol to symbol, and this complicates non-uniform bit-loading. Hence, having the scattered pilots in-place
during the bit-loading process greatly simplifies the bit loading operation. The insertion of continuous pilots, PLC
and nulled regions also results in shift of data subcarriers, but this shift is the same for every symbol, and can easily
be accounted for in the bit loading process.
The CMTS only interleaves data subcarriers and scattered pilots, and therefore only needs information about the
number of data subcarriers and scattered pilots per symbol. In addition, the interleaver does not need to know what
modulation has been applied to an individual data subcarrier. Regardless of modulation scheme, all OFDM symbols
10/29/13
CableLabs
121
CM-SP-PHYv3.1-I01-131029
will have the same number of data subcarriers and scattered pilots, and the modulation pattern of these data
subcarriers may change from symbol to symbol.
7.5.7
IDFT
7.5.7.1
Downstream Transmitter Inverse Discrete Fourier Transform
The CMTS transmitter MUST use the IDFT definition and subcarrier referencing method described in this section.
This section defines the inverse discrete Fourier transform (IDFT) used in the CMTS transmitter for DOCSIS 3.1.
OFDM subcarrier referencing for other definitions such as PLC location, continuous pilots, exclusion bands and bit
loading is also described.
The OFDM signal assembled in the frequency domain consists of 4096 subcarriers for the 4K FFT and 8192
subcarriers for the 8K FFT. The OFDM signal is composed of:
Data subcarriers
Scattered pilots
Continuous pilots
PLC subcarriers
Excluded subcarriers that are zero valued
This signal is described according to the following IDFT equation:

() =
1
=0 () exp(
2 2
) , = 0, 1, , ( 1)
The resulting time domain discrete signal, x(i), is a baseband complex-valued signal, sampled at 204.8 Msamples
per second.
In this definition of the IDFT:
X(0) is the lowest frequency component;
X(N/2) is the DC component or the mean value of the sequence x(i);
X(N-1) is the highest frequency component.
122
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
The IDFT is illustrated in Figure 752.

Frequency
()
k=0
k = N-1

Transform
Time
()
i=0
i = N-1
Figure 752 - Inverse Discrete Fourier Transform
The sample rate in the time domain is 204.8 Msamples/s. Hence, the N samples of the discrete Fourier transform
cover a frequency range of 204.8 MHz. This gives the subcarrier spacing shown in Table 732.
Table 732 - Subcarrier Spacing
IDFT Size N
Carrier Spacing
4096
8192
50 kHz
25 kHz
The maximum channel bandwidth is 192 MHz; this corresponds to 3841 subcarriers in 4K mode and 7681
subcarriers in 8K mode. The active bandwidth of the channel is expected to be 190 MHz; this corresponds to 3801
subcarriers in 4K mode and 7601 subcarriers in 8K mode.
Table 733 describes what the different values of k mean for 4K FFTs and 8K FFTs.
Table 733 - k Definitions for 4K and 8K FFT
Value of k
4K FFT
8K FFT
0
128
148
2048
3948
3968
4095
0
256
296
4096
7896
7936
8191
Description
Lowest frequency subcarrier of the DFT
Lower-end subcarrier of the 192 MHz channel
Lower-end subcarrier of the 190 MHz band
DC component
Upper-end subcarrier of the 190 MHz band
Upper-end subcarrier of the 192 MHz channel
Highest frequency subcarrier of the DFT
The OFDM channel bandwidth can be any value from 24 MHz to 192 MHz; smaller bandwidths than 192 MHz are
achieved by nulling subcarriers X(k) prior to the IDFT. Note that the channel need not be centered at the subcarrier
10/29/13
CableLabs
123
CM-SP-PHYv3.1-I01-131029
k = N/2, although this would be the most logical approach when transmitting a channel with bandwidth less than
192 MHz.
For example, consider transmitting an OFDM signal with a subcarrier spacing of 25 kHz over a 24 MHz channel
with an active bandwidth of 22 MHz. The channel would have 881 active subcarriers, including all edge subcarriers.
The most logical thing to do would be to assign these active subcarriers to:
{(), = 3656, 3657, , 4095, 4096, 4097, , 4535, 4536}
All other subcarriers are nulled for exclusion. This results in a channel that is symmetrically placed around the DC
component of the time domain sequence {(), = 0, 1, , 8191}. However, the 881 active subcarriers may occupy
any other contiguous region of the frequency domain sequence {(), = 0, 1, , 8191}.
7.5.7.2
Subcarrier Referencing
It is necessary to refer to specific OFDM subcarriers for several definitions:

a)
Defining continuous pilot locations
b)
Defining exclusion bands and excluded individual subcarriers
c)
Defining bit loading profiles
Each of these definitions uses the index k of the equation defined in the preceding section to refer to a specific
subcarrier.
The subcarrier index goes from 0 to 4095 for the 4K FFT and from 0 to 8191 for the 8K FFT; each of these
definitions is limited to these subcarrier indices.
The PLC is also defined with reference to k = 0. The OFDM template carried by the PLC defines the subcarrier
index of the lowest frequency subcarrier of the PLC. Hence, once the CM detects the PLC, the CM knows the
location of k = 0. Since the FFT size is also known, it is possible to precisely compute the FFT of the data channel
containing the PLC.
Note that scattered pilot placement is not referenced to k = 0; instead, it is referenced directly to the PLC preamble.
7.5.8
This section describes how cyclic prefixes are inserted and how a window is applied to the output of the IDFT at the
CMTS and how they are handled by the CM.
The addition of a cyclic prefix enables the receiver to overcome the effects of inter-symbol-interference and intercarrier-interference caused by micro-reflections in the channel. Windowing maximizes channel capacity by
sharpening the edges of the spectrum of the OFDM signal. Spectral edges occur at the two ends of the spectrum of
the OFDM symbol, as well as at the ends of internal exclusion bands.
The number of active OFDM subcarriers can be increased by sharpening these spectral edges. However, sharper
spectral edges in the frequency domain imply longer tapered regions in the time domain, resulting in increased
symbol duration and reduction in throughput. Therefore, there is an optimum amount of tapering that maximizes
channel capacity. This optimum is a function of channel bandwidth as well as the number of exclusion bands.
7.5.8.1
Cyclic Prefix Insertion and Windowing
The CMTS MUST follow the procedure described in section 7.4.11.1for cyclic prefix insertion and windowing,
using CMTS specific cyclic prefix and roll-off period values.
The CMTS MUST support cyclic prefix extension and windowing as described in section 7.4.11.1.
The CMTS MUST support the cyclic prefix values defined in Table 734 for both 4K and 8K FFTs.
The CM MUST support the cyclic prefix values listed defined Table 734 for both 4K and 8K FFTs.
124
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Table 734 - Cyclic Prefix (CP) Values

Cyclic Prefix (s)
0.9375
1.25
2.5
3.75
5.0
192
256
512
768
1024
The cyclic prefix (in s) are converted into samples using the sample rate of 204.8 Msamples/s and is an integer
multiple of: 1/64 * 20 s.
The CMTS MUST support the five parameter values specified for this roll-off listed in Table 735.
Table 735 - Roll-Off Prefix (RP) Values
Roll-Off Period (s)
Roll-Off Period Samples (Ncp)
0
0.3125
0.625
0.9375
1.25
0
64
128
192
256
The CMTS MUST NOT allow a configuration in which the RP value is the CP value.
7.5.9
Fidelity Requirements
For the purposes of this specification, the number of occupied CEA channels of an OFDM channel is the occupied
bandwidth of the OFDM channel divided by 6 MHz.
CMTSs capable of generating N-channels of legacy DOCSIS plus NOFDM-channels of OFDM per RF port, for
purposes of the DRFI output electrical requirements, the device is said to be capable of generating Neq-channels per
RF port, where Neq = N + 32*NOFDM "equivalent legacy DOCSIS channels."
An Neq-channel per RF port CMTS MUST comply with all requirements operating with all Neq channels on the RF
port, and MUST comply with all requirements for an Neq-channel per RF port device operating with Neq active
channels on the RF port for all values of Neq less than Neq.
For an OFDM channel there is a) the occupied bandwidth, b) the encompassed spectrum, c) the modulated spectrum,
and d) the number of equivalent legacy DOCSIS channels.
The encompassed spectrum in MHz is 204.8 MHz minus the number of subcarriers in the Band edge Exclusion Subband for the upper and lower band edges (combined) times the subcarrier spacing in MHz. For example, with
subcarrier spacing of 50 kHz and 150 lower band edge subcarriers and 152 upper band edge subcarriers for a total of
302 subcarriers in the two Band edge Exclusion Sub-bands, the encompassed spectrum = 204.8 302*(0.05) =
189.7 MHz. The encompassed spectrum is also equal to the center frequency of the highest frequency modulated
subcarrier minus the center frequency of the lowest frequency modulated subcarrier in an OFDM channel, plus the
subcarrier spacing.
The occupied bandwidth is a multiple of 6 MHz, with a minimum of 24 MHz, and consists of all CEA channels
which include the encompassed spectrum plus taper region shaped by the OFDM channels transmit windowing; the
out-of-band spurious emissions requirements (except for gap channel spurious emissions requirements) apply
outside the occupied spectrum. With a 1 MHz taper region on each band edge of the OFDM channel, shaped by the
transmit windowing function, encompassed spectrum of 189.7 MHz may provide 192 MHz of occupied bandwidth.
The modulated spectrum of an OFDM channel is the encompassed spectrum minus the total spectrum in the Internal
Excluded Sub-bands of the channel, where the total spectrum in the Internal Excluded Sub-bands is equal to the
number of subcarriers in all of the Internal Excluded Sub-bands of the OFDM channel multiplied by the subcarrier
spacing of the OFDM channel. In the previous example, if there are 188 subcarriers total in three Internal Exclusion
10/29/13
CableLabs
125
CM-SP-PHYv3.1-I01-131029
Sub-bands, then the total spectrum in the Internal Excluded Sub-bands (in MHz) is 188*0.05 = 9.4 MHz, and the
modulated spectrum is 189.7 MHz 9.4 MHz = 180.3 MHz.
The number of equivalent active legacy DOCSIS channels in the OFDM channel Neq is the ceiling function applied
to the modulated spectrum divided by 6 MHz. For the example, the number of equivalent legacy DOCSIS channels
in the OFDM channel is ceiling(180.3 MHz / 6 MHz) = 31.
For an Neq-channel per RF port device, the applicable maximum power per channel and spurious emissions
requirements are defined using a value of N* = minimum( 4Neq, ceiling[Neq/4]) for Neq < Neq/4, and N* = Neq
otherwise.
These specifications assume that the CMTS will be terminated with a 75 Ohm load.
7.5.9.1
CMTS Output Electrical Requirements
For OFDM, all modulated subcarriers in an OFDM channel are set to the same average power (except pilots which
are boosted by 6 dB). For purposes of spurious emissions requirements, the "commanded transmit power per
channel" for an equivalent legacy DOCSIS channel is computed as follows:
CMTS power is configured by power per CEA channel and number of occupied CEA channels for each OFDM
channel.
For each OFDM channel, the total power is Power per CEA channel + 10log10(Number of occupied CEA
channels) for that OFDM channel.
CMTS calculates power for data subcarrier and pilots (using total number of non-zero valued (non-excluded)
subcarriers).
CMTS calculates power in 6 MHz containing PLC.
For the spurious emissions requirements, power calculated for the 6 MHz containing the PLC is the commanded
average power of an equivalent DOCSIS legacy channel for that OFDM channel.
A CMTS MUST output an OFDM RF modulated signal with the characteristics defined in Table 737 and Table 7
38. Legacy DOCSIS RF modulated signal characteristics are provided in Section 6.2.22.
The condition for these requirements is all Neq' combined channels, legacy DOCSIS channels and equivalent legacy
DOCSIS channels, commanded to the same average power, except for the Single Channel Active Phase Noise,
Diagnostic Carrier Suppression, OFDM Phase Noise, OFDM Diagnostic Suppression, and power difference
requirements, and except as described for Out-of-Band Noise and Spurious Requirements.
Table 736 - RF Output Electrical Requirements
Parameter
Value
Downstream Lower Edge Band of a CMTS
CMTS MUST support 258 MHz

CMTS SHOULD support 108 MHz
Downstream Upper Edge Band of a CMTS
CMTS MUST support 1218 MHz

CMTS MAY support 1794MHz
Level
Adjustable. See Table 737.
Modulation Type
See Section 7.5.6.1
OFDM channels subcarrier spacing
25 kHz and 50 kHz
Inband Spurious, Distortion, and Noise

528 MHz total occupied bandwidth,
6 MHz gap (Internal Excluded subcarriers)
88 equivalent legacy DOCSIS channels.
See Notes 4, 6
126
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Parameter
Value
For measurements below 600 MHz
-50 dBr Average over center 400 kHz subcarriers within gap
For measurements from 600 MHz to 1002

MHz
For measurements 1002 MHz to 1218 MHz

MER in 192 MHz OFDM channel occupied
bandwidth
528 MHz total occupied bandwidth,
88 equivalent legacy DOCSIS channels. See
Notes 2, 4, 5, 6
48 dB
50 dB
Any single subcarrier. See Note 1

Average over the complete OFDM channel.
See Note 1

MHz
45 dB
47 dB

See Note 1
43 dB
45 dB

See Note 1
Minimal test receiver equalization: See note 7

2 dB relief for above requirements
(e.g., MER > 48 dB becomes MER > 46 dB)
MER in 24 MHz OFDM channel occupied
bandwidth, single OFDM channel only, 24
MHz total occupied spectrum:
See Notes 1, 2, 4, 8
48 dB

MHz
45 dB
43 dB
Phase noise, double sided maximum, Full

power CW signal 1002 MHz or lower
1 kHz - 10 kHz:
10 kHz - 100 kHz:
100 kHz - 1 MHz:
1 MHz - 10 MHz:
10 MHz - 100 MHz:
-48 dBc
-56 dBc
-60 dBc
-54 dBc
-60 dBc
Full power 192 MHz OFDM channel block
10/29/13
CableLabs
127
CM-SP-PHYv3.1-I01-131029
Parameter
Value
with 6 MHz in center as Internal Exclusion

subband + 0 dBc CW in center, with block not
extending beyond 1002 MHz
[CW not processed via FFT]
1 kHz - 10 kHz:
10 kHz - 100 kHz:
-48 dBc
-56 dBc

100 kHz - 1 MHz:
-60 dBc

1 MHz - 10 MHz:
Output Impedance
75 ohms
Output Return Loss (Note 3)
> 14 dB within an active output channel from 88 MHz to 750

MHz
MHz
MHz
> 12 dB in every inactive channel from 54 MHz to 870 MHz
> 10 dB in every inactive channel from 870 MHz to 1218 MHz
Connector
-53dBc
Table Notes:
1. Receiver channel estimation is applied in the test receiver; test receiver does best estimation possible. Transmit windowing is
applied to potentially interfering channel and selected to be sufficient to suppress cross channel interference
2. MER (modulation error ratio) is determined by the cluster variance caused by the transmit waveform at the output of the ideal
receive matched filter. MER includes all discrete spurious, noise, subcarrier leakage, clock lines, synthesizer products,
distortion, and other undesired transmitter products. Phase noise up to 50 kHz of the subcarrier is excluded from inband
specification, to separate the phase noise and inband spurious requirements as much as possible. In measuring MER, record
length or carrier tracking loop bandwidth may be adjusted to exclude low frequency phase noise from the measurement. MER
requirements assume measuring with a calibrated test instrument with its residual MER contribution removed.
3. Frequency ranges are edge-to-edge.
4. Phase noise up to 10 MHz offset is mitigated in test receiver processing or by test equipment (latter using hardline carrier from
modulator, which requires special modulator test port and functionality).
5. Up to 5 subcarriers in one OFDM channel can be excluded from this requirement
6. The measured OFDM channel is allocated 204.8 MHz of spectrum which is free from the other OFDM channel and 24 SCQAM channels which together comprise 528 MHz of occupied spectrum.
7. The estimated channel impulse response used by the test receiver is limited to half of length of smallest transmit cyclic prefix
8. A single subcarrier in the OFDM channel can be excluded from this requirement, no windowing is applied and minimum CP is
selected.
128
CableLabs
10/29/13
7.5.9.1.1
CM-SP-PHYv3.1-I01-131029
Power per Channel for CMTS
A CMTS MUST generate an RF output with power capabilities as defined in Table 737.
The CMTS MUST be capable of adjusting channel RF power on a per channel basis as stated in Table 737.
If the CMTS has independent modulation capability on a per channel basis for legacy DOCSIS channels, then the
CMTS MUST be capable of adjusting power on a per channel basis for the legacy DOCSIS channels, with each
channel independently meeting the power capabilities defined in Table 737.
Table 737 - CMTS Output Power
Parameter
Value
Required power per channel for Neq channels

combined onto a single RF port for
Neq >= Neq /4:
Required power in dBmV per channel
Neq >= Neq /4
60 ceil [3.6*log2(Neq)] dBmV

combined onto a single RF port for Neq < Neq /4:
4 =< Neq < Neq /4
60 ceil [3.6*log2(Neq")] dBmV
Range of commanded transmit power per channel
>= 8 dB below required power level specified below

maintaining full fidelity over the 8 dB range
Range of commanded power per channel; adjusted on

a per channel basis
MUST: 0 dBc to -2 dBc relative to the highest

commanded transmit power per channel, within an 8 dB
absolute window below the highest commanded power.
MAY: required power (in table below) to required
power - 8 dB, independently on each channel.
Commanded power per channel step size
=< 0.2 dB Strictly monotonic
Power difference between any two adjacent channels

in the 108-1218 MHz downstream spectrum (with
commanded power difference removed if channel
power is independently adjustable)
=< 0.5 dB
Power difference between any two non-adjacent

channels in a 48 MHz contiguous bandwidth block
(with commanded power difference removed if
channel power is independently adjustable)
=< 1 dB
Power difference (normalized for bandwidth)

between any two channels OFDM channel blocks or
legacy DOCSIS channels in the 108 - 1218 MHz
downstream spectrum (with commanded power
difference removed if channel power is independently
adjustable)
=< 2 dB
Power per channel absolute accuracy
2 dB
10/29/13
CableLabs
129
CM-SP-PHYv3.1-I01-131029
Parameter
Value
Diagnostic carrier suppression (3 modes)

Mode 1: One channel suppressed
1) >= 50 dB carrier suppression within the occupied

spectrum in any one active channel. This MUST be
accomplished without service impacting discontinuity or
detriment to the unsuppressed channels.
Mode 2: All channels suppressed except one
2) 50 dB carrier suppression within the occupied

spectrum in every active channel except one. The
suppression is not required to be glitchless, and the
remaining unsuppressed active channel is allowed to
operate with increased power such as the total power of
the N active channels combined.
3) 50 dB carrier suppression within the occupied
spectrum in every active channel.
Mode 3: All channels suppressed
In all three modes the output return loss of the

suppressed channel(s) MUST comply with the Output
Return Loss requirements for active channels given in
Table 739 - CMTS Output Requirements.
The total noise and spur requirement is the combination
of noise power from the 50 dBc suppressed channel and
the normal noise and spur requirement for the CMTS
output when operating with all channels unsuppressed.
RF output port muting
>= 73 dB below the unmuted aggregate power of the RF

modulated signal, in every 6 MHz CEA channel from 54
MHz to 1218 MHz.
The specified limit applies with all active channels
commanded to the same transmit power level.
Commanding a reduction in the transmit level of any, or
all but one, of the active channels does not change the
specified limit for measured muted power in 6 MHz.
The output return loss of the output port of the muted
device MUST comply with the Output Return Loss
requirements for inactive channels given in Table 739 CMTS Output Requirements.

combined onto a single RF port for
Neq >= Neq /4:
Neq >= Neq /4
60 ceil [3.6*log2(Neq)] dBmV

combined onto a single RF port for Neq < Neq /4:
Required power in dBmV per channel,
4 =< Neq < Neq /4
60 ceil [3.6*log2(Neq")] dBmV
130
CableLabs
10/29/13
7.5.9.1.2
CM-SP-PHYv3.1-I01-131029
Out-of-Band Noise and Spurious Requirements for CMTS
One of the goals of the DOCSIS DRFI specification is to provide the minimum intended analog channel CNR
protection of 60 dB for systems deploying up to 119 DRFI-compliant QAM channels. A new DOCSIS 3.1 PHY goal
is to provide protection for legacy DOCSIS channels and for high density constellations of OFDM channel blocks if
they are generated from another DRFI-compliant device.
The specification assumes that the transmitted power level of the digital channels will be 6 dB below the peak
envelope power of the visual signal of analog channels, which is the typical condition for 256-QAM transmission. It
is further assumed that the channel lineup will place analog channels at lower frequencies than digital channels, and
in systems deploying DOCSIS 3.1 modulators, analog channels will be placed at center frequencies below 600 MHz.
An adjustment of 10*log10 (6 MHz / 4 MHz) is used to account for the difference in a 6 MHz equivalent digital
channel,, versus an analog channels noise power bandwidth. With the assumptions above, for a 119-6 MHz
equivalent channel system, the specification in item 5 of equates to an analog CNR protection of 60dB.
With all digital channels at the same equivalent power per 6 MHz channel, the specification provides for 58 dB SNR
protection for analog channels below 600 MHz (even with transmissions above 600 MHz) with 192 MHz occupied
spectrum (one full OFDM channel) and 51 dB SNR protection for digital channels below 600 MHz with
transmission of 960 MHz modulated spectrum (160 equivalent legacy DOCSIS channels). The SNR protection
between 600 MHz and 1002 MHz is 55 dB for analog channels operating above a 192 MHz occupied spectrum
generated by a DOCSIS 3.1compliant device, and is 48 dB between 600 MHz and 1002 MHz for digital channels
operating above 960 MHz occupied spectrum generated by a DOCSIS 3.1 compliant device.
Table 738 lists the out-of-band spurious requirements. In cases where the N' combined channels are not
commanded to the same power level, "dBc" denotes decibels relative to the strongest channel among the active
channels. When commanded to the same power level, "dBc" should be interpreted as the average channel power,
averaged over the active channels, to mitigate the variation in channel power across the active channels (see
Table 737), which is allowed with all channels commanded to the same power.
The CMTS modulator MUST satisfy the out-of-band spurious emissions requirements of Table 738 in
measurements below 600 MHz and outside the encompassed spectrum when the active channels are contiguous or
when the ratio of modulated spectrum to gap spectrum within the encompassed spectrum is 4:1 or greater.
The CMTS modulator MUST satisfy the out-of-band spurious emissions requirements of Table 738, with 1 dB
relaxation, in measurements within gaps in modulated spectrum below 600 MHz and within the encompassed
spectrum when the ratio of modulated spectrum to gap spectrum within the encompassed spectrum is 4:1 or greater.
relaxation, when the ratio of modulated spectrum to gap spectrum within the encompassed spectrum is 4:1 or
greater, in measurements with 603 MHz center frequency 999 MHz, outside the encompassed spectrum or in
gap channels within the encompassed spectrum.
relaxation, when the ratio of modulated spectrum to gap spectrum within the encompassed spectrum is 4:1 or
greater, in measurements with 999 MHz < center frequency 1209 MHz, outside the encompassed spectrum or in
gap channels within the encompassed spectrum.
The CMTS modulator MUST satisfy the out-of band spurious emissions requirements of Table 738, in addition to
contributions from theoretical transmit windowing, with permissible configurations of lower edge and upper edge
subband exclusions of at least 1 MHz each, FFT Size, cyclic prefix length (Ncp) and windowing roll-off period (Nrp)
values. Recommendations for configuration parameters are provided in Appendix V. The test limit for determining
compliance to the spurious emissions requirements is the power sum of the spurious emissions requirements taken in
accordance with Table 738; and the contributions from the theoretical transmit windowing for the configured
transmissions.
When the Neq combined active channels are not contiguous, and the ratio of modulated spectrum to gap spectrum
within the encompassed spectrum is 4:1 or greater, the spurious emissions requirements are determined by summing
the spurious emissions power allowed in a given measurement bandwidth by each of the contiguous sub-blocks
among the occupied spectrum. In the gap channels within the encompassed spectrum and below 600 MHz there is a
1 dB relaxation in the spurious emissions requirements, so that within the encompassed spectrum the spurious
emissions requirements (in absolute power) are 26% higher power in the measurement band determined by the
10/29/13
CableLabs
131
CM-SP-PHYv3.1-I01-131029
summing of the contiguous sub-blocks spurious emissions requirements. In all channels above 600 MHz there is a 3
dB relaxation in the spurious emissions requirements, so that the spurious emissions requirements (in absolute
power) are double the power in the measurement band determined by the summing of the contiguous sub-blocks
spurious emissions requirements. The following three paragraphs provide the details of the spurious emissions
requirements for non-contiguous channel operation outside the encompassed spectrum; within the encompassed
spectrum the same details apply except there is an additional 1 dB allowance below 600 MHz; and 3 dB allowance
is applied above 600 MHz for all channels.
The full set of Neq channels is referred to throughout this specification as the modulated channels or the active
channels. However, for purposes of determining the spurious emissions requirements for non-contiguous transmitted
channels, each separate contiguous sub-block of channels within the active channels is identified, and the number of
channels in each contiguous sub-block is denoted as Neqi, for i = 1 to K, where K is the number of contiguous subblocks. Therefore, Neq = i=1 to k Neqi. Note that K = 1 when and only when the entire set of active channels is
contiguous. Also note that an isolated transmit channel, i.e., a transmit channel with empty adjacent channels, is
described by Ni = 1 and constitutes a sub-block of one contiguous channel. Any number of the "contiguous subblocks" may have such an isolated transmit channel; if each active channel was an isolated channel, then K = N.
When Neq >= Neq/4, Table 738 is used for determining the noise and spurious power requirements for each
contiguous sub-block, even if the sub-block contains fewer than Neq/4 active channels. When Neq < Neq/4, Table 7
38 is used for determining the noise and spurious power requirements for each contiguous sub-block. Thus, the noise
and spurious power requirements for all contiguous sub-blocks of transmitted channels are determined from
Table 738, where the applicable table is determined by Neq being greater than or equal to Neq/4, or not. The noise
and spurious power requirements for the ith contiguous sub-block of transmitted channels is determined from
Table 738 using the value Ni for the "number of active channels combined per RF port", and using "dBc" relative
to the highest commanded power level of a 6 MHz equivalent channel among all the active channels, and not just the
highest commanded power level in the ith contiguous sub-block, in cases where the Neq combined channels are not
commanded to the same power. The noise and spurious emissions power in each measurement band, including
harmonics, from all K contiguous sub-blocks, is summed (absolute power, NOT in dB) to determine the composite
noise floor for the non-contiguous channel transmission condition.
For the measurement channels adjacent to a contiguous sub-block of channels, the spurious emissions requirements
from the non-adjacent sub-blocks are divided on an equal "per Hz" basis for the narrow and wide adjacent
measurement bands. For a measurement channel wedged between two contiguous sub-blocks, adjacent to each, the
measurement channel is divided into three measurement bands, one wider in the middle and two narrower bands
each abutting one of the adjacent transmit channels. The wideband spurious and noise requirement is split into two
parts, on an equal "per Hz" basis, to generate the allowed contribution of power to the middle band and to the
farthest narrowband. The ceiling function is applied to the resulting sum of noise and spurious emissions, per Note 1
of Table 738 to produce a requirement of dB resolution.
Items 1 through 4 list the requirements in channels adjacent to the commanded channels.
Item 5 lists the requirements in all other channels further from the commanded channels. Some of these "other"
channels are allowed to be excluded from meeting the Item 5 specification. All the exclusions, such as 2nd and 3rd
harmonics of the commanded channel, are fully identified in the table.
Item 6 lists the requirements on the 2N' 2nd harmonic channels and the 3N' 3rd harmonic channels.
Table 738 - CMTS Output Out-of-Band Noise and Spurious Emissions Requirements
for
[ ,
,
Band
], < /
/
, Adjusted Number of Active Channels
Combined per RF Port
Requirement (in dBc)
For N* = 1, 2, 3, 4: -58;
1
132
Adjacent channel up to 750 kHz from channel block edge
For N* 5: 10*log10 [10-58/10 +(0.75/6)*(10-65/10 +

(N*-2)*10-73/10)]
CableLabs
10/29/13
for
CM-SP-PHYv3.1-I01-131029
[ ,
,
Band
], < /
/
, Adjusted Number of Active Channels
Combined per RF Port
Requirement (in dBc)
For N* = 1: -62;
2
Adjacent channel (750 kHz from channel block edge to 6

MHz from channel block edge)
Next-adjacent channel (6 MHz from channel block edge to

12 MHz from channel block edge)
Third-adjacent channel (12 MHz from channel block edge to

18 MHz from channel block edge)
For N* 2: 10*log10 [10-62/10 +(5.25/6)*(10-65/10

+(N*-2)*10-73/10)]
10*log10 [10-65/10 +( N*-1)*10-73/10]
For N* = 1: -73;
For N* = 2: -70;
For N* = 3: -67;
For N* = 4: -65;
For N* = 5: -64.5;
For N* = 6, 7: -64;
For N* 8: -73 + 10*log10 (N*)
Noise in other channels (47 MHz to 1218 MHz) Measured in

each 6 MHz channel excluding the following:
a) Desired channel(s)
b) 1st, 2nd, and 3rd adjacent channels (see Items 1, 2, 3, 4
in this table)
c) Channels coinciding with 2nd and 3rd harmonics (see
Item 6 in this table)
For N* = 1:
For N* = 2:
For N* = 3:
For N* = 4:
-73;
-70;
-68;
-67;
For N* 5: -73 + 10*log10 (N*)
In each of 2N contiguous 6 MHz channels or in each of 3N

contiguous 6 MHz channels coinciding with 2nd harmonic
and with 3rd harmonic components respectively (up to 1218
MHz)
-73 + 10*log10(N*) dBc, or -63, whichever is

greater
Lower out of band noise in the band of 5 MHz to 47 MHz

Measured in 6 MHz channel bandwidth
-50 + 10*log10(N*)
Higher out of band noise in the band of 1218 MHz to 3000

MHz
Measured in 6 MHz channel bandwidth
For N* 8: -55 + 10*log10(N*)
For N* > 8: -60 + 10*log10(N*)
Table Notes
1. All equations are Ceiling(Power, 0.5) dBc. Use "Ceiling(2*Power) / 2" to get 0.5 steps from ceiling functions that
return only integer values. For example Ceiling(-63.9, 0.5) = -63.5 dBc.
2. Add 3 dB relaxation to the values specified above for noise and spurious emissions requirements in all channels with
603 MHz center frequency 999 MHz. For example -73 dBc becomes -70 dBc.
Add 5 dB relaxation to the values specified above for noise and spurious emissions requirements in all channels with
999 MHz < center frequency 1209 MHz . For example -73 dBc becomes -68 dBc.
3. Add 1 dB relaxation to the values specified above for noise and spurious emissions requirements in gap channels
with center frequency below 600 MHz. For example -73 dBc becomes -72 dBc.
7.5.10 CMTS Transmitter Output Requirements

CMTSs capable of generating N-channels of legacy DOCSIS plus NOFDM-channels of OFDM per RF port, for
purposes of the output electrical requirements, the device is said to be capable of generating Neq-channels per RF
port, where Neq = N + 32*NOFDM "equivalent legacy DOCSIS channels."
An Neq-channel per RF port device MUST comply with all requirements operating with all Neq channels on the RF
port, and MUST comply with all requirements for an Neq-channel per RF port device operating with Neq active
channels on the RF port for all values of Neq less than Neq.
10/29/13
CableLabs
133
CM-SP-PHYv3.1-I01-131029
The number of equivalent legacy DOCSIS channels in the OFDM channel is the ceiling function applied to the
modulated spectrum divided by 6 MHz. For the example, the number of equivalent legacy DOCSIS channels in the
OFDM channel is ceiling(180.3 MHz / 6 MHz) = 31.
For an Neq-channel per RF port device, the applicable maximum power per channel and spurious emissions
requirements are defined using a value of N* = minimum( 4Neq, ceiling[Neq/4]) for Neq < Neq/4, and N* = Neq
otherwise.
These specifications assume that the CMTS will be terminated with a 75 ohm load.
7.5.10.1 CMTS Output Electrical Requirements
A CMTS MUST output an RF modulated signal with the characteristics defined in Table 739. The condition for
these requirements is the entire OFDM block commanded to a constant power spectral density, except for Phase
Noise, Diagnostic Carrier Suppression, and power difference requirements (Table 737), and except as described for
Out-of-Band Noise and Spurious Requirements listed in Section 7.5.9.
When the CMTS generates 54 MHz (nine or more equivalent channels) on a single RF output port, it MUST
provide for independent selection of center frequency when the ratio of active bandwidth to excluded bandwidth in
the encompassed spectrum is at least 2:1, and with each channel independently meeting the requirements of
Table 739 except for spurious emissions requirements defined in Section 7.5.9.
When the CMTS generates 54 MHz (nine or more equivalent channels) on a single RF output port it MUST meet
the requirements of Table 739 when the ratio of active bandwidth to excluded bandwidth in the encompassed
spectrum is at least 4:1.
Table 739 - CMTS Output Requirements
Parameter
Value
Lower Band Edge
Refer to Section 7.2
Upper Band Edge
Refer to Section 7.2.
Signal Type
OFDM
Single FFT Block Bandwidth
192 MHz
Minimum Active Signal Bandwidth
24 MHz
Subcarrier Spacing / OFDM Symbol Rate 25 kHz / 40 usec

FFT duration
50 kHz / 20 susec
FFT Size
50 kHz: 4096 (4K FFT)

25 kHz: 8192 (8K FFT)
Maximum Number of Subcarriers per

FFT
4K: 3840
8K: 7680
Number of Data Subcarriers per FFT
4K: 3801 number of continuous pilot tones

8K: 7601 number of continuous pilot tones
Continuous Pilot Tones
Active bandwidth > 48 MHz: 1 of every 128 subcarriers

Active bandwidth 48 MHz: 1 of every 64 subcarriers, 128 maximum
Scattered Pilot Tones
One out of every 4K FFT: 128 subcarriers, staggered by 1

8K FFT: 64 subcarriers, staggered by 2128 subcarriers
Samples
sec
192
0.9375
256
1.25
512
2.5
768
3.75
1024
5.0
Cyclic Prefix Options
* sampling rate of 204.8 MHz
134
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Parameter
OFMD OFDM Shaping

Windowing Options
Value
Raised cosine (Tukey) absorbed by CP

Samples
sec
0
0
64
0.3125
128
0.625
192
0.9375
256
1.25
* sampling rate of 204.8 MHz
Level
Adjustable, see Table 737.
Inband Spurious, Distortion, and Noise
See Table 736
Inband Spurious and Noise
See Table 736
Out of Band Spurious and Noise
See Table 738Table 736
In band Phase Noise
See Table 736
Allowable Degradation
Output Impedance
1.5 dB
75 ohms
Output Return Loss
See Table 736
Connector
F connector per [SCTE 02]
7.5.10.1.1 Independence of individual channel within the multiple channels on a single RF port
A potential use of a CMTS is to provide a universal platform that can be used for high-speed data services or for
video services. For this reason, it is essential that interleaver depth be set on a per channel basis to provide a suitable
transmission format for either video or data as needed in normal operation. Any N-channel block of a CMTS MUST
be configurable with at least two different interleaver depths, using any of the interleaver depths defined in Section
7.5.6.1. Although not as critical as per-channel interleaver depth control, there are strong benefits for the operator if
the CMTS is provided with the ability to set RF power, center frequency, and modulation type on a per-channel
basis.
1.
A multiple-channel CMTS MUST be configurable with at least two different legacy interleaver depths among
the legacy channels on an RF output port in addition to each OFDM channel which is configurable
independently.
2.
A multiple-channel CMTS MUST provide for 3 modes of carrier suppression of RF power for diagnostic and
test purposes, see Table 737 for mode descriptions and carrier RF power suppression level.
3.
A multiple-channel CMTS MAY provide for independent adjustment of RF power in a per channel basis for
legacy channels with each RF carrier independently meeting the requirements defined in Table 737
4.
A multiple-channel CMTS MAY provide for independent selection of center frequency on a per channel basis,
thus providing for non-contiguous channel frequency assignment with each channel independently meeting the
requirements in Table 739. A multiple-channel CMTS capable of generating nine or more channels on a single
RF output port MUST provide for independent selection of center frequency with the ratio of number of active
channels to gap channels in the encompassed spectrum being at least 2:1, and with each channel independently
meeting the requirements in Table 739 except for spurious emissions. A multiple-channel CMTS capable of
generating nine or more channels on a single RF output port MUST meet the requirements of Table 739 when
10/29/13
CableLabs
135
CM-SP-PHYv3.1-I01-131029
the ratio of number of active channels to gap channels in the encompassed spectrum is at least 4:1. (A ratio of
number of active channels to gap channels of at least 4:1 provides that at least 80% of the encompassed
spectrum contains active channels, and the number of gap channels is at most 20% of the encompassed
spectrum.)
5.
A multiple-channel CMTS MAY provide for independent selection of modulation order, either 64-QAM or
256-QAM, on a per channel basis for legacy channels, with each channel independently meeting the
requirements in Table 739.
6.
A CMTS MUST provide a test mode of operation, for out-of-service testing, configured for N channels but
generating one-CW-per-channel, one channel at a time at the center frequency of the selected channel; all other
combined channels are suppressed. One purpose for this test mode is to support one method for testing the
phase noise requirements of Table 739. As such, the CMTS generation of the CW test tone SHOULD exercise
the signal generation chain to the fullest extent practicable, in such manner as to exhibit phase noise
characteristics typical of actual operational performance; for example, repeated selection of a constellation
symbol with power close to the constellation RMS level would seemingly exercise much of the modulation and
up-conversion chain in a realistic manner. The CMTS test mode MUST be capable of generating the CW tone
over the full range of Center Frequency in Table 739.
7.
A CMTS MUST provide a test mode of operation, for out-of-service testing, generating one-CW-per-channel,
at the center frequency of the selected channel, with all other N 1 of the combined channels active and
containing valid data modulation at operational power levels. One purpose for this test mode is to support one
method for testing the phase noise requirements of Table 739. As such, the generation of the CW test tone
SHOULD exercise the signal generation chain to the fullest extent practicable, in such manner as to exhibit
phase noise characteristics typical of actual operational performance. For example, a repeated selection of a
constellation symbol, with power close to the constellation RMS level, would seemingly exercise much of the
modulation and upconversion chain in a realistic manner. For this test mode, it is acceptable that all channels
operate at the same average power, including each of the N 1 channels in valid operation, and the single
channel with a CW tone at its center frequency. The test mode MUST be capable of generating the CW tone
over the full range of Center Frequency in Table 739.
8.
A CMTS MUST be capable of glitchless reconfiguration over a range of active channels from ceiling[7*Neq/8]
to Neq. Channels which are undergoing configuration changes are referred to as the "changed channels." The
channels which are active and are not being reconfigured are referred to as the "continuous channels".
Glitchless reconfiguration consists of any of the following actions while introducing no discontinuity or
detriment to the continuous channels, where the modulator is operating in a valid DOCSIS 3.1-required mode
both before and after the reconfiguration with an active number of channels staying in the range
{ceiling[7*Neq/8], Neq}: adding and/or deleting one or more channels, and/or moving some channels to new RF
carrier frequencies, and/or changing the interleaver depth, modulation, power level, or frequency on one or
more channels. Any change in the modulation characteristics (power level, modulation density, interleaver
parameters, center frequency) of a channel excuses that channel from being required to operate in a glitchless
manner. For example, changing the power per channel of a given channel means that channel is not considered
a continuous channel for the purposes of the glitchless modulation requirements. Glitchless operation is not
required when Neq is changed, even if no reconfigurations accompany the change in Neq.
7.5.11 Cable Modem Receiver Input Requirements

The CM MUST be able to accept any range of OFDM subcarriers defined between Lower Frequency Boundary and
Upper Frequency Boundary simultaneously. Active subcarrier frequencies, loading, and other OFDM characteristics
are described by OFDM configuration settings and CM exclusion bands and profile definition. The OFDM signals
and CM interfaces will have the characteristics and limitations defined in Table 740.
Table 740 - Electrical Input to CM
Parameter
Lower Frequency Boundary
136
Value
MUST support 258 MHz

SHOULD support 108 MHz
Note: applies if fumax is 85 MHz or less
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Parameter
Value
Upper Frequency Boundary
MUST support 1218 MHz

SHOULD support 1794 MHz
Frequency Boundary Assignment

Granularity
25 kHz 8K FFT
50 kHz 4K FFT
Signal Type
OFDM
Single FFT Block Bandwidth
192 MHz
Minimum Active OFDM Bandwidth
24 MHz
Number of FFT Blocks
CM MUST support minimum of 2 FFT Blocks AND

24 SC-QAM Channels
Subcarrier Spacing/FFT Duration
25 kHz / 40 s
50 kHz / 20 s
Modulation Type
MUST support
QPSK, 16-QAM, 64-QAM, 128-QAM, 256-QAM, 512-QAM,
1024-QAM, 2048-QAM, 4096-QAM
MAY support
8192-QAM, 16384-QAM
Variable Bit Loading
MUST support with subcarrier granularity

MUST support zero bit loaded and zero valued subcarriers
Total Input Power
< 40 dBmV, 54 MHz 1.794 GHz

* Assuming negligible power outside this range
Level Range (24 MHz min occupied BW)

Equivalent PSD to SC-QAM of
-15 dBmV to + 15 dBmV per 6 MHz
-9 dBmV/24 MHz
-21 dBmV/24 MHz
Maximum average power of any 24 MHz

input to the CM from 54 MHz to 1218 MHz
OR
From 258 MHz to 1.794 GHz
Let X = Average power of lowest power 24 MHz BW for

demodulation
Additional Demodulated Bandwidth, Bremond :
Min [X + 10 + 10*log(Bremond/24) ; 21+ 10*log(Bremond/24)]
Additional Non-Demodulated Bandwidth, Bno-demod :
Min [X + 10 + 10*log(Bno-demod/24) ; 26+ 10*log(Bno-demod/24)]
For up to 12 MHz of occupied bandwidth (analog, OOB, QAM,
OFDM)
Min [X + 10 + 10*log(Bno-demod/24) ; 21+ 10*log(Bno-demod/24)]
for all remaining bandwidth
Level range does not imply anything about BER performance or
capability vs. QAM. CM BER performance is separately described.
Input Impedance
10/29/13
75 ohms
CableLabs
137
CM-SP-PHYv3.1-I01-131029
Parameter
Value
Input Return Loss
> 6 dB (258 MHz 1218 MHz)

> 6 dB (108 MHz 1218 MHz)
Note: Applies when lower frequency boundary is 108 MHz
> 6 dB (258 MHz 1.794 GHz)
Note: Applies when upper frequency boundary is 1.794 GHz
Connector
7.5.12 Cable Modem Receiver Capabilities

The required level for CM downstream post-FEC error rate is defined as less than or equal to 10-6 PER (packet error
ratio) with 1500 byte Ethernet packets. This section describes the conditions at which the CM is required to meet
this error rate.
7.5.12.1 CM Error Rate Performance in AWGN Channel
Implementation loss of the CM MUST be such that the CM achieves the required error rate when operating at a
CNR as shown in Table 741, under input load and channel conditions as follows:
Any valid transmit combination (frequency, subcarrier clock frequency, transmit window, cyclic prefix,
pilot, PLC, subcarrier exclusions, interleaving depth, multiple modulation profile configuration, etc.) as
defined in this spec.
P6AVG (the measured channel power divided by number of occupied CEA channels) 15 dBmV.
Up to fully loaded spectrum of 54 - 1218 MHz, including up to 48 analog channels.
Power in (both above and below) 4 adjacent 6 MHz channels P6AVG+3 dB.
Power in any 6 MHz channel over the spectrum P6AVG+6 dB.
Peak envelope power in any analog channel over the spectrum P6AVG+6 dB
Average power per channel across spectrum P6AVG+3 dB.
OFDM channel phase noise as in CMTS spec.
No other artifacts (reflections, burst noise, tilt, etc.).

Table 741 - CM Minimum CNR Performance in AWGN Channel
1,2
1,2
Constellation
CNR (dB)
Up to 1 GHz
CNR (dB)
1 GHz to 1.2 GHz
Min P6AVG dBmV
4096
41.0
41.5
-6
2048
37.0
37.5
-9
1024
34.0
34.0
-12
512
30.5
30.5
-12
256
27.0
27.0
-15
128
24.0
24.0
-15
64
21.0
21.0
-15
15.0
15.0
-15
16
Table Notes:
1. CNR is defined here as total signal power in occupied bandwidth divided by total noise in occupied bandwidth
2. Channel CNR is adjusted to the required level by measuring the source inband noise including phase noise
component and adding the required delta noise from an external AWGN generator
3. Applicable to an OFDM channel with 192 MHz of occupied spectrum
138
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
7.5.13 Physical Layer Link Channel (PLC)

This section contains the description of the Physical layer Link Channel (PLC) that the CMTS follows during the
construction of the PLC.
The aim of the PLC is for the CMTS to convey to the CM the physical properties of the OFDM channel. In a blind
acquisition, that is, in an acquisition without prior knowledge of the physical parameters of the channel, the CM first
acquires the PLC, and from this extract the parameters needed to acquire the complete OFDM channel.
7.5.13.1 PLC Placement
The CMTS MUST transmit a PLC for every downstream OFDM channel.
The CMTS MUST place this PLC in a 1 MHz grid, that is, the center frequency of the lowest frequency subcarrier
of the PLC has to be an integer when the frequency is measured in units of MHz.
7.5.13.2 PLC Structure
The CMTS MUST place the PLC so that it occupies the same set of contiguous subcarriers in every OFDM symbol.
The CMTS MUST place 8 OFDM subcarriers in the PLC of every OFDM symbol when using 4K FFT OFDM (i.e.,
a subcarrier spacing of 50 kHz).
The CMTS MUST place 16 OFDM subcarriers in the PLC of every OFDM symbol when using 8K FFT OFDM
(i.e., a subcarrier spacing of 25 kHz).
Table 742 - PLC components
DFT size
Subcarrier spacing
Number of
PLC subcarriers (Np)
4096
8192
50 kHz
25 kHz
8
16
The CMTS MUST use a 16-QAM constellation for the PLC subcarriers.
The CMTS MUST construct the PLC as 8 symbols of preamble followed by 120 symbols of data subcarriers, as
shown in Figure 753.
Frequency
NP
Preamble
8 symbols
PLC Data
Preamble
Data
120 symbols
Time
Figure 753 - Structure of the PLC
10/29/13
CableLabs
139
CM-SP-PHYv3.1-I01-131029
The CMTS MUST place the PLC at the center of a 6 MHz of active frequency range. For 4K FFT OFDM, this 6
MHz channel, in the increasing order of frequency, will consist of 56 subcarriers followed by the 8 PLC subcarriers
followed by another 56 subcarriers. For 8K FFT OFDM, this 6 MHz channel, in the increasing order of frequency,
will consist of 112 subcarriers followed by the 16 PLC subcarriers followed by another 112 subcarriers.
The CMTS MUST NOT insert any exclusion zones or excluded subcarriers within this 6 MHz band that contains the
PLC, including the PLC.
The CMTS MUST insert 8 continuous pilots in this 6 MHz bandwidth, 4 on each side of the PLC, as defined in the
section on downstream pilots.
The CMTS MUST interleave the PLC subcarriers on their own, as described in the section on "PLC Interleaving".
The CMTS MUST NOT interleave the PLC preamble.
The CMTS MUST synchronize the scattered pilot pattern to the PLC preamble as defined in section 7.5.15 such that
in the OFDM symbol that follows the last symbol of the preamble sequence, the subcarrier next to the highestfrequency subcarrier in the PLC is a scattered pilot.
The CMTS MUST NOT insert any scattered pilots or continuous pilots within the PLC frequency band.
The CMTS MUST synchronize the downstream data randomizer to the PLC preamble as described in the section on
"Downstream Data randomization". That is, the CMTS must initialize the downstream randomizer just before the
lowest frequency data subcarrier of the first OFDM symbol following the preamble.
The CMTS MUST synchronize the downstream PLC randomizer to the PLC preamble as described in the section on
"Downstream PLC randomization". That is, the CMTS MUST initialize the downstream PLC randomizer just before
the lowest frequency PLC subcarrier of the first OFDM symbol following the preamble.
The CMTS MUST place the 6 MHz bandwidth containing the PLC within the active bandwidth of the OFDM
channel.
Two possible locations for the PLC channel are illustrated in the example of Figure 754. In this example there are
three contiguous OFDM spectral bands in the 192 MHz channel, of width 22, 12 and 5 MHz. There are two
exclusion zones between these. The spectrum outside these bands is also excluded.
It is not necessary to place the PLC in the largest contiguous spectral segment of the OFDM channel. The 6 MHz
channel containing the PLC at its center may be anywhere provided it contains 6 MHz of spectrum without any
excluded subcarriers. In the example given the one possible location for the PLC channel is in the 12 MHz wide
segment.
Since the downstream channel will contain a minimum of 22 MHz of contiguous OFDM spectrum, there will always
be a spectral band to place the PLC. It may be noted that it not necessary to place the PLC at the center of the 22
MHz bandwidth.
The CMTS is expected to place the PLC in part of the spectrum that is less susceptible to noise and interference.
140
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Time
6 MHz
Possible PLC Location
22 MHz
Exclusion Zone 1
6 MHz
PLC Possible Location
12 MHz
Exclusion Zone 2
5 MHz
Frequency
Figure 754 - Examples of PLC placement
10/29/13
CableLabs
141
CM-SP-PHYv3.1-I01-131029
The CMTS MUST generate the PLC as shown in Figure 755:
Map Bytes
to a
Bit Stream
LDPC
Encode
(384,288)
Map Bit
Stream to
nibbles
Randomize
the nibbles
Map each nibble to

QAM-16
Constellation point
Interleave
blocks of
96 nibbles
Time Multiplex
with BPSK
Preamble
Time and Frequency interleaved

data and NCP subcarriers
Continuous and Scattered Pilots
Form
OFDM
Symbol in
Frequency
Domain &
Compute
IDFT
Excluded subcarriers
Figure 755 - Physical Layer Operations for Forming the PLC Subcarriers
7.5.13.3 PLC Preamble Modulation

The CMTS MUST modulate the subcarriers in the PLC preamble using binary phase-shift keying (BPSK), as
described in this section.
For 4K FFT, the preamble size is 8 subcarriers. Thus, an array of size (8 x 8) is defined as follows:
Table 743 - PLC preamble for 4K FFT
Symbol 1
Symbol 2
Symbol 3
Symbol 4
Symbol 5
Symbol 6
Symbol 7
Symbol 8
Subcarrier 1
Subcarrier 2
Subcarrier 3
Subcarrier 4
Subcarrier 5
Subcarrier 6
Subcarrier 7
Subcarrier 8
142
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Table 744 - PLC preamble for 8K FFT
For 8K FFT, the preamble size is 16 subcarriers. Thus, an array of size (16 x 8) is defined as follows:
Symbol 1
Symbol 2
Symbol 3
Symbol 4
Symbol 5
Symbol 6
Symbol 7
Symbol 8
Subcarrier 1
Subcarrier 2
Subcarrier 3
Subcarrier 4
Subcarrier 5
Subcarrier 6
Subcarrier 7
Subcarrier 8
Subcarrier 9
Subcarrier 10
Subcarrier 11
Subcarrier 12
Subcarrier 13
Subcarrier 14
Subcarrier 15
Subcarrier 16
The CMTS MUST map each of the above binary bits to a BPSK constellation point in the complex plane using the
following transformation:
0 (1 + j0)
1 (-1 + j0)
7.5.13.4 PHY Parameters Carried by the PLC
The PLC carries two sets of PHY parameters from the CMTS to cable modems: the Downstream Profile Descriptor
and the OFDM Channel Descriptor. Contents of each of these descriptors are described in [DOCSIS MULPIv3.1].
This section contains only a brief description of the physical layer parameters carried by the PLC. For formatting
and other details, reference is made to the MULPI specification.
The inverse discrete Fourier transform that defines the OFDM signal at the CMTS is given by the following
equation:
x(i) =
N1
k=0 X(k) exp(
j2ik 2
N
); for i = 0, 1, , N 1
(1)
The sampling rate in the previous equation is 204.8 Msamples/s and the value of N is either 4096 or 8192. The
CMTS MUST specify this value of N via the PLC.
The CMTS MUST define, via the PLC, the frequency of the subcarrier X(0) in equation (1) as a 32-bit positive
integer in units of Hz.
The PLC subcarriers constitute a set of contiguous subcarriers given by:
{X(k), k = L, L + 1, , L + NP 1}
(2)
The CMTS MUST define the value of L to define the location of the PLC within an OFDM channel.
The CMTS MUST define the locations of the continuous pilots, excluding the eight predefined ones, via the PLC.
The CMTS MUST define the locations of excluded subcarriers via the PLC.
10/29/13
CableLabs
143
CM-SP-PHYv3.1-I01-131029
The CMTS MUST define the bit loading profile for all 4096 or 8192 subcarriers of equation (1), excluding
continuous pilots and excluded subcarriers, via the PLC.
The CMTS MUST use the indices k of equation (1) to specify the locations of subcarriers in all of the above
definitions.
In addition to above, the CMTS MUST define the following physical parameters of the OFDM channel:
Cyclic prefix length (five possible settings)
Roll-off (five possible settings)
Time interleaver depth (any integer from 0 to 32.
Modulation of the NCP (QPSK, QAM-16 or QAM-64)
7.5.13.5 Mapping of Bytes to a Bit Stream

The CMTS MUST convert the stream of bytes received by the PLC into a stream of bits, MSB first, as illustrated in
Figure 756.
Time
Byte 2
Byte 1
a23
a22
a21
a20
a19
a18
a17
a16
a15
a14
a13
a12
a11
a10
a9
a8
Byte 0
a7
a6
a5
a4
a3
a2
a1
a0
LSB
Byte
Stream
MSB
a20 a19 a18 a17 a16 a15 a14 a13 a12 a11 a10 a9 a8 a7 a6 a5 a4 a3 a2 a1 a0
Bit Stream
Time
Figure 756 - Mapping Bytes into a Bit Stream for FEC Encoding
7.5.13.6 Forward Error Correction code for the PLC

The CMTS MUST encode the PLC data using (384,288) puncturing LDPC encoder, see Section 7.4.4.3 for the
definition of puncturing encoder.
The puncturing encoder uses the same mother encoder for fine ranging FEC (Section7.4.15.2.3), that is the rate 3/5
(480,288) LDPC encoder listed by Table 714.
Denote the information bits sent to the mother code encoder by ( a0 ,, a 287 ) and let the encoder output being
(a0 ,, a 287 , b288 ,, b479 ) , where b288 ,, b479 are parity-check bits. The coordinates to be deleted by the
puncturing step are:
Period 1: 48 consecutive coordinates a 48 ,, a95
Period 2: 48 consecutive coordinates b384 ,, b431
144
CableLabs
10/29/13
Note:
CM-SP-PHYv3.1-I01-131029
Also see Figure 770.
The puncturing is described in Figure 757.
Mother code
Encoder
a0 a1 a287
288
a0 a 287 b288 b479
a0 a1 a47
a96 a97 a287 b288b289 b383
b432 b479
48
(1st period)
48
( 2nd period)
a0 a1 a 47 a96 a97 a 287 b288 b383b432 b469
240
144
Figure 757 - Puncturing Encoder for the PLC FEC
7.5.13.7 Block Interleaving of the PLC Subcarriers

The preceding section shows 240 data bits entering the LDPC encoder and 384 encoded bits exiting the LDPC
encoder. This sequence is in effect is time-reversed order. The time-ordered sequence takes the form shown in the
figure below.
The CMTS MUST map these 384 data bits into 96 4-bit nibbles {ui , i = 0,1, , 95} as described below using
Figure 758 before interleaving.
Time
a0 a1 a2 a3 a4 a5 a6 a7 a47 a96 a97 a287 b288 b289 b383 b432 b469
MSB
LSB
u0,0=a0
u0,1=a1
u0,2=a2
u0,3=a3
u1,0=a4
u1,1=a5
u1,2=a6
u1,3=a7
u0
u1
u95,0=b466
u95,1=b467
u95,2=b468
u95,3=b469
u95
Figure 758 - Mapping Encoded Bit Stream into a Stream of Nibbles
The CMTS MUST interleave this 96-nibble sequence {u0 u1 u2 u95 } as described below.
For 4K FFT, the CMTS MUST use an (8x12) array. The CMTS MUST write the values along the rows of this
two-dimensional array, as shown in Figure 759.
10/29/13
CableLabs
145
CM-SP-PHYv3.1-I01-131029
12
u84
u85
u95
Frequency
8
u12
u0
u13
u23
u11
u1
Time
Figure 759 - Block Interleaving of PLC Subcarriers for 4K FFT
The CMTS MUST then read this two-dimensional array along vertical columns to form the two-dimensional
sequence {vt,f , t = 0, 1, , 11 and f = 0, 1, , 7}. This operation is mathematically represented as:
vt,f = ut+12f
The CMTS MUST map each of the 8-point sequence given below to the 8 successive PLC subcarriers of an OFDM
symbol after randomization described in the next section.
Vt = {vt.f , f = 0, 1, , 7}
for 12 successive OFDM symbols t =0, 1, , 11
Therefore each FEC codeword will occupy the PLC segment of twelve successive 4K FFT OFDM symbols. There
will be ten such codewords in an 128-symbol PLC frame, including the 8-symbol preamble.
The CMTS MUST map ten complete FEC codewords into one 4K FFT OFDM frame.
For 8K FFT, the CMTS MUST use a (16 x 6) array. The CMTS MUST write the values ui along the rows of this
two-dimensional array, as shown Figure 760.
146
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
6
u90
u91
u95
Frequency
16
u6
u11
u7
u0
u1
u5
Time
Figure 760 - Block Interleaving of PLC Subcarriers for 8K FFT
The CMTS MUST then read this two-dimensional array along vertical columns to form the two-dimensional
sequence {vt,f , t = 0, 1, , 5 and f = 0, 1, , 15}. This operation is mathematically represented as:
vt,f = ut+6f
The CMTS MUST map each of the 16-point sequence given below to the 16 successive PLC subcarriers of an
OFDM symbol after randomization described in the next section.
Vt = {vt.f , f = 0, 1, , 15} for 6 successive OFDM symbols t =0, 1, , 5
Therefore, each FEC codeword will occupy the PLC segment of six successive 8K FFT OFDM symbols. There will
be twenty such codewords in an128-symbol PLC frame, including the 8-symbol preamble.
The CMTS MUST map twenty complete FEC codewords into one 8K FFT OFDM frame.
7.5.13.8 Randomizing the PLC Subcarriers
The CMTS MUST randomize the sequence {vt,f , t = 0, 1, , 5 and f = 0, 1, , 15} using a copy of the linear
feedback shift register in GF[212] used for randomizing the data subcarriers. This is shown in Figure 761.
10/29/13
CableLabs
147
CM-SP-PHYv3.1-I01-131029
12-bit
D1
D0
11
12-bit
Figure 761 - Linear Feedback Shift Register for PLC Randomization
The LFSR is defined by the following polynomial in GF[212].

x 2 + x + 11
The GF[212] is defined through polynomial algebra modulo the polynomial:

12 + 6 + 4 + + 1
This LFSR is initialized to the hexadecimal numbers given below:

D0 = "4A7"
D1 = "B4C"
The CMTS MUST initialize the LFSR with the above two 12-bit numbers at the beginning of the first OFDM
symbol following the PLC preamble. The CMTS MUST clock the LFSR once after randomizing one PLC
subcarrier. The CMTS MUST randomize each subcarrier through an exclusive-OR operation of the 4 bits
representing the subcarrier (vt,f ) with the four LSBs of register D0.
The first subcarrier to be randomized is the lowest frequency subcarrier of the PLC in the OFDM symbol
immediately after the preamble. This will be randomized using the four LSBs of the initialized D0, namely 0x7. The
LFSR will be clocked once after randomizing each PLC subcarrier of the OFDM symbol. After randomizing the
highest frequency PLC subcarrier of an OFDM symbol the CMTS MUST clock the LFSR before randomizing the
lowest frequency PLC subcarrier in the next OFDM symbol.
The CMTS MUST use the bit ordering given below to perform randomization. The four LSBs of D0 are defined as
the coefficients of {3 2 1 0 } of the Galois field polynomial representing D0. The LSB is defined as the
coefficient of 0 of the polynomial representing D0. The ordering of the four bits representing the subcarrier is
defined with reference to Figure 758. Assume that the FEC block shown in Figure 758 is the first FEC block in
the PLC frame. Then, since the location of the first nibble does not change as a result of interleaving:
v0,0 = {a0 a1 a2 a3 }
Then the randomization operation (i.e., exclusive-OR with 0x7) is given by:
{y0 , y1 , y2 , y3 } = {a0 + 0, a1 + 1, a 2 + 1, a 3 + 1}
The addition operations in the above equation are defined in GF[2], that is, these are bit-wise exclusive-OR
operations. The LFSR is clocked once before randomizing the next nibble v0,1 .
The CMTS MUST NOT randomize the PLC preamble.

7.5.13.9 Mapping to 16-QAM Subcarriers
The CMTS MUST map each randomized nibble {y0 y1 y2 y3 } into a complex number using the 16-QAM constellation
mapping shown in Figure A4.
148
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
The CMTS MUST multiply the real and imaginary parts by 1/10 to ensure that mean-square value of the QAM
constellation is unity.
7.5.13.10 PLC Timestamp Reference Point
The PLC subcarriers following the preamble may contain a timestamp.
The CMTS MUST define this timestamp with reference to the first OFDM symbol following the preamble if such a
timestamp exists. This OFDM symbol is indicated by an arrow in Figure 762.
Frequency
Preamble
Time
Figure 762 - Time - Frequency Plane Representation of PLC Timestamp Synchronization
Time domain version of the OFDM symbol is shown in Figure 763. The inverse discrete Fourier transform of the
symbol of Figure 762 results in the set of 4096 or 8192 samples occupying the FFT duration shown. After this the
CMTS will introduce a configurable cyclic prefix (CP), window the symbol and overlap successive symbols in the
time domain.
The CMTS MUST use the time of the first sample of the FFT duration as the timestamp.
To clarify this further, individual time domain samples are also shown in Figure 763. (This is for illustration only;
actual samples are complex-valued.) The sample rate is 204.8 Msamples/s. The dotted arrow points to the first
sample of the FFT symbol duration.
10/29/13
CableLabs
149
CM-SP-PHYv3.1-I01-131029
FFT Duration (useful symbol period)

N + NCP
Sample spacing is 1/204.8 s
Time
Time
Figure 763 - Time Domain Representation of PLC Timestamp Synchronization

7.5.14.1 Mapping of Bytes to Bits
Each NCP consists of three bytes as defined in Section 8.3.4. The first byte (Byte 0) contains the profile identifier as
the four MSBs and four control bits as the four LSBs. The other two bytes (Byte 1 and Byte 2) contain the start
pointer.
The CMTS MUST map the three NCP bytes into 24-bit serial bit stream {a23 a22 a 0 } for the purpose of LDPC
encoding, as shown in Figure 764. Note that the LDPC encoder is also defined using the same bit pattern
{a23 a22 a 0 }.
150
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Byte 2
Byte 1
a23
a22
a21
a20
a19
a18
a17
a16
a15
a14
a13
a12
a11
a10
a9
a8
Byte 0
a7
a6
a5
a4
a3
a2
a1
a0
LSB
MSB
Bit Stream
a23 a22 a21 a20 a19 a18 a17 a16 a15 a14 a13 a12 a11 a10 a9 a8 a7 a6 a5 a4 a3 a2 a1 a0
Time
Figure 764 - Mapping NCP Bytes into a Bit Stream for FEC Encoding
7.5.14.2 Forward Error Correction code for the NCP

The CMTS MUST encode the 24 information bits of a Next Codeword Pointer using (48, 24) shortening and
puncturing LDPC encoder, see Section 7.4.13.2.3 for the definition of shortening and puncturing encoder.
The shortening puncturing encoder uses the same mother encoder for initial ranging FEC (Section 7.4.4.3), that is
the rate 1/2 (160, 80) LDPC encoder listed by Table 713.
Denote the information bits sent to the mother code encoder by
(a 0 , , a 79 , b80 , , b159 ) , where b80 , , b159
(a 0 , , a 79 )
and let the encoder output being
are parity-check bits. Then the shortening and puncturing steps
can be described as follows, also see Figure 7-79:

The shortening step fills 0 to 56 consecutive coordinate starting at position 24, i.e., let a 24
= a 25 = = a 79 = 0 .
The rest 24 bits i.e. a 0 , , a 23 , are NCP information data.

The coordinates to be deleted by the puncturing step are:
Period 1: 24 consecutive coordinates
b80 , , b103
b112 , , b127
b144 , , b159
10/29/13
CableLabs
151
CM-SP-PHYv3.1-I01-131029
a0 a1 a22 a23 00
000
a0 a1 a22 a23
56
24
24 information bits
Mother
Code
Encoder
b128 b143
a0 a1 a23
b104 b111
16
(2nd period)
80
Shortening
Fill
24
(1st period)
a0 a23 b80 b159
a0 a23 000b80 b159b80
a0 a1 a23 b104b105 b111b128b129 b143

16
( 3rd period)
24
24
Puncturing
Figure 765 - Shortening and Puncturing Encoder for the NCP FEC
7.5.14.3 Mapping LDPC Encoded Bits into to OFDM Subcarriers

The LDPC encoder outputs a stream of 48 bits:
{b143 b142 b128 b111 b110 b104 a23 a22 a 0 }
The NCP QAM constellation can be a member of the set {QPSK, QAM-16, QAM-64}.
For QAM-64 the CMTS MUST map the LDPC encoded bits into eight 6-bit QAM constellation points as defined
below:
y0,0 y0,1 y0,2 y0,3 y0,4 y0,5 = {a5 a4 a3 a2 a1 a0 }
y1,0 y1,1 y1,2 y1,3 y1,4 y1,5 = {a11 a10 a9 a8 a7 a6 }
y7,0 y7,1 y7,2 y7,3 y7,4 y7,5 = {b143 b142 b141 b140 b139 b138 }
The mapping of these 6-bit integers to points in the complex plane is given by the figure below. Hexadecimal
notation has been used to represent the 6-bit numbers{yi,0 yi,1 yi,2 yi,3 yi,4 yi,5 }.
152
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
20
22
2A
28
08
0A
02
00
21
23
2B
29
09
0B
03
01
25
27
2F
2D
0D
0F
07
05
24
26
2E
2C
0C
0E
06
04
-1
34
36
3E
3C
1C
1E
16
14
-3
35
37
3F
3D
1D
1F
17
15
-5
31
33
3B
39
19
1B
13
11
-7
30
32
3A
38
18
1A
12
10
-7
-5
-3
-1
Figure 766 - 64-QAM Constellation Mapping of {, , , , , , }
For QAM-16 the CMTS MUST map the LDPC encoded bits into twelve 4-bit QAM constellation points as defined
below:
y0,0 y0,1 y0,2 y0,3 = {a3 a2 a1 a0 }
y1,0 y1,1 y1,2 y1,3 = {a7 a6 a5 a4 }
y11,0 y11,1 y11,2 y11,3 = {b143 b142 b141 b140 }
notation has been used to represent the 4-bit numbers{yi,0 yi,1 yi,2 yi,3 }.
10/29/13
CableLabs
153
CM-SP-PHYv3.1-I01-131029
-1
-3
-1
-3
Figure 767 - 16QAM Constellation Mapping of {, , , , }
For QPSK the CMTS MUST map the LDPC encoded bits into twenty four 2-bit QAM constellation points as
defined below:
y0,0 y0,1 = {a1 a0 }
y1,0 y1,1 = {a3 a2 }
y23,0 y23,1 = {b143 b142 }
notation has been used to represent the 2-bit numbers{yi,0 yi,1 }.
-1
-1
Figure 768 - QPSK Constellation Mapping of {, ,}
7.5.14.4 Placement of NCP Subcarriers

The CMTS MUST place the NCP subcarriers beginning from the frequency location of the highest frequency active
data subcarrier of the OFDM symbol, and going downwards along active data subcarriers of the OFDM symbols
before they are time and frequency interleaved.
154
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Therefore the first subcarrier of the first NCP occupies the frequency location of the highest frequency active data
subcarrier of the OFDM symbol. The term active data subcarrier is used to indicate a subcarrier that is neither
excluded and that is neither a continuous pilot nor a scattered pilot. This highest frequency active subcarrier may not
occur at the same frequency in every symbol owing to the presence of scattered pilots.
The OFDM symbol, prior to time and frequency interleaving at the CMTS, will have subcarriers assigned to be
scattered pilot placeholders. Furthermore, the NCP profile may indicate subcarriers that are to be zero-bit-loaded.
The CMTS MUST skip both of these types of subcarriers during the placement of NCP subcarriers.
7.5.14.5 Randomization and Interleaving of NCP Subcarriers
The CMTS MUST randomize the NCP subcarriers using the algorithm applied to the data subcarriers, described in
section on data subcarrier randomization.
The CMTS MUST time and frequency interleave the NCP subcarriers using the algorithm applied to data
subcarriers and this is described in section interleaving.
7.5.15 Downstream Pilot Patterns
Downstream pilots are subcarriers modulated by the CMTS with a defined modulation pattern that is known to all
the CMs in the system to allow interoperability.
There are two types of pilots: continuous and scattered. Continuous pilots occur at fixed frequencies in every
symbol. Scattered pilots occur at different frequency locations in different symbols. Each of these pilot types for
DOCSIS 3.1 is defined in the following sections.
7.5.15.1 Scattered Pilots
The main purpose of scattered pilots is the estimation of the channel frequency response for the purpose of
equalization. There are two scattered pilot patterns, one for 4K FFT and one for 8K FFT. Although these pilots
occur at different frequency locations in different OFDM symbols, the patterns repeat after every 128 OFDM
symbols; in other words, the scattered pilot pattern has a periodicity of 128 OFDM symbols along the time
dimension.
7.5.15.1.1 Scattered Pilot Pattern for 4K FFT
The CMTS MUST create scattered pilots for 4K FFTs in the manner described in this section.
Figure 769 shows the 4K FFT scattered pilot pattern for OFDM transmissions.
The scattered pilot pattern is synchronized to the PLC as shown in Figure 769. The first OFDM symbol after the
PLC preamble has a scattered pilot in the subcarrier just after the highest frequency subcarrier of the PLC. Two such
scattered pilots that are synchronized to the PLC preamble are marked as red circles in Figure 772.
The remainder of the scattered pilot pattern is linked to the scattered pilot synchronized to the PLC preamble, using
the following rules:
1.
In each symbol scattered pilots are placed every 128 subcarriers.
2.
From symbol to symbol, scattered pilots are shifted by one subcarrier position in the increasing direction of
the frequency axis. This will result in scattered pilots placed in the exclusion band and in the PLC band.
3.
Scattered pilots are nulled in the exclusion bands; all the subcarriers in the exclusion bands are zero-valued
subcarriers.
4.
Scattered pilots are nulled when these coincide with nulled subcarriers; all nulled subcarriers are zerovalued subcarriers.
5.
In the PLC, normal PLC signals (i.e., PLC data or the PLC preamble) are transmitted instead of scattered
pilots. The CMTS MUST NOT transmit scattered pilots in the PLC band.
10/29/13
CableLabs
155
CM-SP-PHYv3.1-I01-131029
Time
8
Symbols
PLC
Data
128
Symbols
PLC
Preamble
PLC
Preamble
PLC
Data
Frequency
128
subcarriers
Figure 769 - 4K FFT Downstream Pilot Pattern
There are 8 preamble symbols in the PLC; for 4K FFT, there are 8 PLC subcarriers in each symbol.
156
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Mathematically, the scattered pilot pattern for a 4K FFT is defined as follows. Let a subcarrier (depicted in red in the
above figure just after the PLC preamble) be referred to as x(m,n), where:
m is the frequency index
n is the time index (i.e., the OFDM symbol number)
The scattered pilots in the 128 symbols following (and including symbol n) are given by:
Symbol n:
Symbol (n+1):
Symbol (n+2):
Symbol (n+127):
x(n, m128i), for all non-negative integers i
x(n+1, m128i + 1) , for all non-negative integers i
Each of the above locations is a scattered pilot, provided that it does not fall on a continuous pilot, on the PLC, on an
exclusion zone or on a excluded subcarrier. If the scattered pilot coincides with a continuous pilot it is treated as a
continuous pilot and not as a scattered pilot.
This pattern repeats every 128 symbols. That is, symbol (128+n) has the same scattered pilot pattern as symbol n.
7.5.15.1.2 Scattered Pilot Pattern for 8K FFT
The CMTS MUST create scattered pilots for 8K FFTs in the manner described in this section.
Figure 770 shows a scattered pilot pattern that may be used for OFDM transmissions employing 8K FFT. This is
used here for explanation purposes only and to help with the derivation of the scattered pilot pattern actually used in
8K FFT OFDM transmissions depicted in Figure 771.
10/29/13
CableLabs
157
CM-SP-PHYv3.1-I01-131029
Time
128
Symbols
PLC
Data
PLC
Prea
mble
PLC
Prea
mble
PLC
Data
Frequency
2 subcarriers
(50 kHz)
128
subcarrier
Figure 770 - A Downstream Scattered Pilot Pattern for 8K FFT (for Explanation Purposes Only)
The scattered pilot pattern is synchronized to the PLC as shown in Figure 769. The first OFDM symbol after the
PLC preamble has a scattered pilot in the subcarrier just after the highest frequency subcarrier of the PLC. Two such
scattered pilots that are synchronized to the PLC preamble are marked as red circles in Figure 770.
In the case of an 8K FFT, pilots are stepped by two subcarriers from one OFDM symbol to the next. Since the pilot
spacing along the frequency axis is 128, this results in a pilot periodicity of 64 in the time dimension. When Figure
158
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
769 and Figure 770 are compared, it is clear that the periodicity is half for the 8K scattered pilot pattern.
However, because an 8K symbol is twice as long as a 4K symbol, the scattered pilot periodicity in terms of actual
time is approximately the same for both the 4K and 8K FFTs. This allows channel estimates for 8K FFTs to be
obtained in approximately the same amount of time as for the 4K FFT. However, scattered pilots for 8K FFTs do not
cover all subcarrier locations and hence intermediate channel estimates have to be obtained through interpolation.
Noise can also be estimated using scattered pilots, and again, the noise at subcarrier locations not covered by
scattered pilots in the 8K FFT can be obtained through interpolation. Note that this interpolation operation could fail
in the presence of narrowband ingress; interpolation could also be problematic when there are excluded subcarriers.
To overcome these interpolation problems, the entire 8K scattered pilot location can be shifted by one subcarrier
location after 64 subcarriers, as illustrated in Figure 771. This may be treated as the interlacing of two identical
scattered pilot patterns. The set of purple scattered pilots are shifted one subcarrier space in relation to the set of
green scattered pilots. As a result the scattered pilots cover all subcarrier locations; noise at every subcarrier location
can be estimated without interpolation. Note that periodicity of the 8K FFT scattered pilot pattern is now 128,
not 64.
10/29/13
CableLabs
159
CM-SP-PHYv3.1-I01-131029
Time
128
Symbols
PLC
Data
PLC
Prea
mble
PLC
Prea
mble
PLC
Data
Frequency
128
subcarriers
Figure 771 - 8K FFT Downstream Scattered Pilot Pattern
Mathematically, the scattered pilot pattern for an 8K FFT is defined as follows. Let the subcarrier (depicted in red in
Figure 771 just after the PLC preamble) be referred to as (, ) where:
m is the frequency index
n is the time index (i.e., the OFDM symbol number)
160
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
The scattered pilots in the first 64 symbols following and including symbol n are given by:
Symbol n:
Symbol (n+1):
Symbol (n+2):
Symbol (n+63):
x(n, m 128i), for all non-negative integers i
x(n + 1, m 128i + 2), for all non-negative integers i

The scattered pilot sequence of the next 64 symbols is the same as above, but with a single subcarrier
shift in the frequency dimension.
Symbol (n+64):
Symbol (n+65):
Symbol (n+66):
Symbol (n+127):

Each of the above locations is a scattered pilot, provided that it does not fall on a continuous pilot, on the PLC, on an
exclusion zone or on a nulled subcarrier. If the scattered pilot coincides with a continuous pilot it is treated as a
continuous pilot and not as a scattered pilot.
This pattern repeats every 128 symbols. That is, symbol (128+n) has the same scattered pilot pattern as symbol n.
7.5.15.2 Continuous Pilots
Continuous pilots occur at the same frequency location in all symbols and are used for receiver synchronization.
Placement of continuous pilots is determined in two ways:
a)
Predefined continuous pilot placement around the PLC
b)
Continuous pilot placement defined via PLC messages
Note that continuous and scattered pilots can overlap; the amount of overlap, in terms of number of carriers, changes
from symbol to symbol. Overlapping pilots are treated as continuous pilots.
7.5.15.2.1 Predefined Continuous Pilots around the PLC
As discussed in Section 7.5.13.1, the PLC is placed at the center of a 6 MHz spectral region. Four pairs of
predefined continuous pilots are placed symmetrically around the PLC as shown in Figure 772. The spacing
between each pilot pair and the PLC are different to prevent all pilots from being impacted at the same time by echo
or interference.
10/29/13
CableLabs
161
CM-SP-PHYv3.1-I01-131029
Time
d3
d4
d2
d1
PLC
d1
d2
d3
d4
Frequency
Figure 772 - Placement of Predefined Continuous Pilots Around the PLC
The locations of the continuous pilots are defined with reference to the edges of the PLC band. Hence, once the PLC
has been detected, these continuous pilots also become known to the receiver.
Table 745 provides the values of d1, d2, d3, and d4, measured in number of subcarriers from the PLC edge. That is,
dx is absolute value of the difference between the index of the continuous pilot and the index of the PLC subcarrier
at the PLC edge nearest to the continuous pilot. The index of a subcarrier is the integer k of the IDFT definition
given in Section 7.5.2.8. For example, let the lowest frequency subcarrier of the PLC have the IDFT index k equal to
972. Then according to Table 736 for the 4K FFT mode the continuous pilot nearest to this lowest frequency PLC
subcarrier will have the IDFT index k of (972-15)=957. The index k of the highest frequency PLC subcarrier of this
OFDM channel is 979. Hence continuous pilot that is nearest upper frequency edge of the PL has an index k of 994.
The table provides the number of subcarriers from the edge of the PLC to the placement of the pilot for the two FFT
sizes. For each distance (dx) defined in Table 745, the CMTS MUST place two pilots: one dx subcarriers above and
one dx subcarriers below the edge of the PLC band.
Table 745 - Subcarrier Distances for Placement of Predefined Pilots
162
d1
d2
d3
d4
4K FFT
PLC 8 subcarriers
15
24
35
47
8K FFT
PLC 16 subcarriers
30
48
70
94
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
7.5.15.2.2 Continuous Pilot Placement Defined by PLC Message

The CMTS MUST define a set of continuous pilots distributed as uniformly as possible over the entire OFDM
spectrum in addition to the predefined continuous pilots described in the preceding section.
The CMTS MUST ensure that there are no isolated active OFDM spectral regions that are not covered by
continuous pilots.
It is not practical to predefine the locations of this set of continuous pilots because of exclusion bands and excluded
subcarriers.
The CMTS MUST provide the continuous pilot placement definition via the PLC in accordance with messaging
formats contained in the MULPI specification.
The CMTS MUST adhere to the rules given below for the definition of this set of continuous pilot locations
conveyed to the CM via PLC messaging. It is noted that these rules do not apply to the eight predefined pilots.
The CMTS MUST place the continuous pilots generated using these rules in every OFDM symbol, in addition to the
eight predefined continuous pilots.
The CMTS MUST obtain the value of NCP using the following formula:
= min max 8,

1906
, 120
(1)
In this equation refers to frequency in Hz of the highest frequency active subcarrier and refers to
frequency in Hz of the lowest frequency active subcarrier of the OFDM channel. It is observed that the number of
continuous pilots is linearly proportional to the frequency range of the OFDM channel. It may also be observed that
the minimum number of continuous pilots defined using the PLC cannot be less than 8, and the maximum number of
continuous pilots defined using the PLC cannot exceed 120. Therefore, the total number of continuous pilots,
including the predefined ones, will be in the range 16 to 128, both inclusive.
The value of M in equation (1) is kept as a parameter that can be adjusted by the CMTS. Nevertheless, the CMTS
MUST ensure that M is in the range given by the following equation:
120 48
(2)
The typical value proposed for M is 48.

The CMTS MUST use the algorithm given below for defining the frequencies for the location of these continuous
pilots.
Step 1:
Merge all the subcarriers between and eliminating the following:
1) Exclusion bands
2) 6 MHz band containing the PLC

3) Known regions of interference, e.g., LTE
4) Known poor subcarrier locations, e.g., CTB/CSO
Let the merged frequency band be defined as the frequency range [0, _ ].
Step 2:
Define a set of NCP frequencies using the following equation:

=
_
2
; for = 0, 1, , 1
(3)
This yields a set of uniformly spaced NCP frequencies:

_
10/29/13
3_
2
, , _
_
2
CableLabs
(4)
163
CM-SP-PHYv3.1-I01-131029
Step 3:
Map the set of frequencies given above to the nearest subcarrier locations in the merged spectrum. This will give a
set of NCP approximately uniformly spaced subcarriers in the merged domain.
Step 4:
De-merge the merged spectrum through the inverse of the operations through which the merged spectrum was
obtained in step 1.
Step 5:
If any continuous pilot is within 1 MHz of a spectral edge, move this inwards (but avoiding subcarrier locations
impacted by interferences like CSO/CTB) so that every continuous pilot is at least 1 MHz away from a spectral
edge. This is to prevent continuous pilots from being impacted by external interferences. If the width of the spectral
region does not allow the continuous pilot to be moved 1 MHz from the edge then the continuous pilot has to be
placed at the center of the spectral band.
Step 6:
Identify any spectral regions containing active subcarriers (separated from other parts of the spectrum by exclusion
bands on each side) that do not have any continuous pilots. Introduce an additional continuous pilot at the center of
every such isolated active spectral region.
In the unlikely event that the inclusion of these extra pilots results in the total number of continuous pilots defined
by PLC exceeding 120, return to step 1 and re-do the calculations after decrementing the value of NCP by one.
Step 7:
Test for periodicity in the continuous pilot pattern and disturb periodicity, if any, through the perturbation of
continuous pilot locations using a suitable algorithm. A simple procedure would be to introduce a random
perturbation of up to 5 subcarrier locations around each continuous pilot location, but avoiding subcarrier locations
impacted by interferences like CSO/CTB.
The CMTS MUST transmit this continuous pilot pattern to the CMs in the system using the PLC.
7.5.15.3 Pilot Modulation
For both continuous and scattered pilots, the CMTS MUST modulate these subcarriers as described in the following
section.
Continuous and scattered pilots are BPSK modulated using a pseudo-random sequence. This pseudo-random
sequence is generated using a 13-bit linear feedback shift register, shown in Figure 773 with polynomial
(x^13+x^12+x^11+x^8+1).
This linear feedback shift register is initialized to all ones at the k=0 index of the 4K or 8K discrete Fourier
transform defining the OFDM signal (refer to Section 7.5.7). It is then clocked after every subcarrier of the FFT. If
the subcarrier is a pilot (scattered or continuous), then the BPSK modulation for that subcarrier is taken from the
linear feedback shift register output.
wk
D
Figure 773 - 13-Bit Linear Feedback Shift Register for the Pilot Modulation Pseudo-Random Sequence
164
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
For example, let the output of the linear feedback shift register be wk. The BPSK modulation used for the pilot
would be:
= 0: = 1 + 0
= 1: = 1 + 0
7.5.15.4 Pilot Boosting
The CMTS MUST multiply the real and imaginary components of continuous and scattered pilots by a real-valued
number such that the amplitude of the continuous and scattered pilots is twice the root-mean-square value of the
amplitude of other subcarriers of the OFDM symbol; That is, continuous and scattered pilots are boosted by
approximately 6 dB with reference to other subcarriers.
10/29/13
CableLabs
165
CM-SP-PHYv3.1-I01-131029
8 PHYMAC CONVERGENCE
8.1
Scope
This specification defines the electrical characteristics and signal processing operations for a cable modem (CM) and
Cable Modem Termination System (CMTS). It is the intent of this specification to define an interoperable CM and
CMTS such that any implementation of a CM can work with any CMTS. It is not the intent of this specification to
imply any specific implementation.
This section describes CM and CMTS requirements for the convergence logical layer between the MAC and PHY
layers for OFDM downstream channels and OFDMA upstream channels. The primary roles of the convergence
layer are to map DOCSIS MAC frames into codewords and to map codewords into minislots for transmission from
the cable modem to the CMTS. Contents of the Next Codeword Pointer (NCP) message and PHY Link Channel
(PLC) are also defined in Section 8.
8.1 Scope
Describes the structure of the DOCSIS 3.1 physical layer (PHY) specification PHY-MAC Convergence chapter.
8.2 Upstream
Defines cable modem and CMTS requirements for convergence layer functions for data transmitted from the
cable modem to the CMTS using minislot and subslot subcarrier structures. Section 8.2 is organized into
subsections as summarized below:
8.2.1 Profiles
Describes the structure and contents of upstream profiles, cable modem requirements for interpreting profile
configuration commands and configuring upstream transmissions according to the configured profile, and
CMTS requirements for communicating upstream profile configuration information to cable modems.
8.2.2 Minislots
Describes the structure and rules for upstream profiles.
8.2.3 Subslots
Describes the structure and pilot pattern options for subslots. Defines cable modem requirements for mapping
data into subslots and CMTS requirements for recovering data from subslots.
8.3 Downstream
Defines cable modem and CMTS requirements for convergence layer functions for data transmitted from the
CMTS to cable modems. Section 8.3 is organized into subsections as summarized below:
8.3.1 Operation
Describes the DOCSIS 3.1 PHY-MAC convergence layer downstream operation including identification of
functional components, description of data flow, and introduction to downstream profiles.
8.3.2 MAC Frame to Codewords
Describes the procedure for creating a FEC codeword from MAC frames. Defines CMTS requirements for
mapping MAC frames to codewords.
8.3.3 Subcarrier Numbering Conventions
Describes the convention for numbering subcarriers in the downstream OFDM channel.
Describes the purpose for and structure of Next Codeword Pointer (NCP) message blocks and the placement of
NCP message blocks in the downstream OFDM channel. Identifies the PHY parameters communicated from
the CMTS to cable modems in the NCP. Defines CMTS requirements for including NCP message blocks in
OFDM downstream channels.
166
CableLabs
10/29/13
8.2
Upstream
8.2.1
Profiles
CM-SP-PHYv3.1-I01-131029
Upstream profiles are comprised of multiple minislots, and are characterized by bit loading and pilot pattern. Bit
loading and pilot patterns can vary between minislots within the profile. The bit loading and pilot pattern assignment
of minislots can also vary between profiles. An upstream profile maps to an Interval Usage Code defined in an
Upstream Channel Descriptor Message.
Different FEC codeword sizes may use portions of a single minislot. The use case for this is as follows: With a 17
kb grant, there needs to be a long codeword to cover the first 16200 bits and a 1 kb codeword to cover the rest of the
bits. The first long codeword can land in the middle of a minislot. In this situation, it does not make sense to require
a constant codeword size per profile, as the profile needs to cover a group of minislots.
FEC codewords can cross minislot and frame boundaries.
8.2.2
Minislots
Minislots are defined by a size in terms of the number of symbols and number of subcarriers. They include data
carried on data subcarriers, pilots carried on pilot subcarriers and complementary pilots that can carry data but at a
lower modulation order.
8.2.2.1
Minislot Parameters
The CMTS MUST define minislot parameters according to Table 81. The CMTS communicates minislot definition
to the CM in UCD messages as defined in [DOCSIS MULPIv3.1].
The CM MUST use the minislot structure defined by the UCD messages received from the CMTS.
The CMTS MUST be capable of receiving minislots structured according to Table 81.
The CMTS MUST apply any subcarrier exclusions to the entire channel independent of upstream profile
assignment.
Table 81 - Minislot Parameters
Parameter
K
Number of symbol periods per frame
Minimum
Value
6
Maximum
Value
For BW > 72 MHz
18 for 20 s FFT duration
Recommended or
Typical Value
N/A

For 48 < BW < 72 MHz
For BW < 48 MHz
Q
Number of subcarriers per minislot
10/29/13
8 for 20 s FFT duration 16 for 40 s FFT duration
CableLabs
167
CM-SP-PHYv3.1-I01-131029
Parameter
Data bitloading
Minimum
Value
QPSK
Maximum
Value
CMTS Mandatory:
1024-QAM
Recommended or
Typical Value
Plant-dependent
CMTS Optional:
2048-QAM
4096-QAM
CM Mandatory:
4096-QAM
Complementary-pilot bitloading
BPSK
256-QAM
Modulation order for primary pilots
BPSK
BPSK
Table Notes
1.
Data bitloading is constant within a minislot, excepting pilots and complementary pilots.
2.
The bitloading of complementary pilots within a minislot is constant.
8.2.2.2
Minislot Structure
This specification defines several minislot structures with pilot patterns as described in Section 7.4.16.
8.2.2.2.1
Number of OFDM Symbols per Minislot
The CMTS MUST follow the rules listed below for the range of the frame size in number of symbols (K):
K is configurable between 6 (minimal value) and one of the following values:

With 20 s FFT duration (4K FFT)
8.2.2.2.2
Kmax = 18 for BW > 72 MHz
Kmax = 24 for 48 MHz < BW < 72 MHz
Kmax = 36 for BW < 48 MHz
With 40 s FFT duration (8K FFT)
Kmax = 9 for BW > 72 MHz
Kmax = 12 for 48 MHz < BW < 72 MHz
Kmax = 18 for BW < 48 MHz

Number of subcarriers per minislot
The CMTS signals the number of subcarriers per minislot to the CM in the UCD.
The CMTS MUST use 16-subcarrier minislots when the subcarrier spacing is 25 kHz.
The CMTS MUST use 8-subcarrier minislots when the subcarrier spacing is 50 kHz.
8.2.2.3
Modulation of Data Subcarriers
With the exception of pilots and complementary pilots, bit loading is constant within a minislot.
The CMTS MUST use the same modulation order for all data subcarriers in a minislot.
168
CableLabs
10/29/13
8.2.2.4
CM-SP-PHYv3.1-I01-131029
Location of Pilots
A set of pilot patterns is defined from which the CMTS or operator can select to match the frequency response of the
network. Pilot patterns are described in Section 7.4.16.
8.2.2.5
Modulation Order of Pilots
The CMTS MUST use BPSK modulation for pilots.

8.2.2.6
Location of Complementary Pilots
The CMTS MUST place complementary pilots as defined by the chosen minislot pattern.
8.2.2.7
Modulation Order of Complementary Pilots
The CMTS MUST use a modulation order equal to (data modulation order 4) for complementary pilots.
The CMTS MUST use a minimum modulation order of BPSK for complementary pilots.
8.2.2.8
Pilot Overhead
Pilot overhead is dependent on the chosen minislot pattern. Capacity and pilot overhead vary with the length of the
minislot (number of symbols) and with the number of subcarriers (8 or 16). Minimum capacity and largest pilot
overhead occur with the shortest minislot length (8 symbols).
8.2.2.9
Mapping Minislots to Subcarriers
The CMTS MUST construct minislots using only contiguous subcarriers. There is no subcarrier exclusion within a
minislot.
8.2.2.10 Ordering of Data Bits within a Minislot
With the exception of initial ranging transmissions, the CM MUST fill minislots as follows: prior to interleaving,
data would be filled across all symbol periods, subcarrier by subcarrier, transmitted symbol period by symbol
period, with complementary pilots filled inline. The fill order is illustrated in Figure 81.
Figure 81 - Minislot Data Bit Ordering
8.2.2.11 Modulation Order Variability

Different minislots are allowed to have different pilot patterns: pilot patterns are assigned at minislot granularity.
Different minislots are allowed to have different bit loading: bit loading is assigned at minislot granularity. This
allows bit loading to vary across the spectrum.
8.2.3
Subslots
This section specifies the subslot within the upstream data frame.
10/29/13
CableLabs
169
CM-SP-PHYv3.1-I01-131029
8.2.3.1
Subslot Structure
The minislot can be subdivided along time into multiple subslots to provide multiple transmission opportunities for
BW requests. The subslots fit within minislot boundary and can have leftover symbols in the end along time axis.
Leftover symbols are not a part of any subslot and are unused (zero valued subcarriers) when the minislot is granted
to IUCs 1 or 2. Gaps between subslots within a single mini-slot are not permitted.
Figure 82 - Subslot Structure
The subslot length is fixed at 2 symbols with mini-slots that employ 16 subcarriers, and 4 symbols with mini-slots
that employ 8 subcarriers.
8.2.3.2
Data mapping among subslots
The data mapping within a subslot is as follows: the mapping starts from the lowest subcarriers index, and lowest
symbol index, and first along symbol time index, and then goes up on subcarriers index. The pilots are skipped
during data mapping.
Data mapping to subcarriers is implemented without time or frequency interleaving.
Figure 83 and Figure 84 illustrate the mapping process.
Figure 83 - Data Mapping for a 4x8 Subslot
170
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Figure 84 - Data Mapping for a 2x16 Subslot
8.3
Downstream
8.3.1
Operation
An example implementation of the downstream convergence layer for DOCSIS 3.1 and its association with the
stages before and after it is shown in Figure 85. This block diagram is intended to demonstrate functionality; while
it represents one style of implementation, there are no requirements that an implementation needs to adhere directly
to this example.
10/29/13
CableLabs
171
CM-SP-PHYv3.1-I01-131029
Figure 85 - Downstream Convergence Layer Block Diagram
The operation of the downstream can be split between the forwarding plane and the control plane. The forwarding
plane contains the data packets that are destined to the user. The control plane carries MAC management messages
and other types of control messages.
The forwarding engine in the CMTS forwards packets to a DOCSIS MAC Domain. The MAC Domain performs
QoS functions such as hierarchical QoS, per-user rate shaping, aggregate per-channel rate shaping, and aggregate
per-bonding-group rate shaping. The MAC engine can also hold packets in a buffer as part of the DOCSIS Light
Sleep mode.
A profile is a list of modulations that are used for the subcarriers within an OFDM channel. The downstream can use
different profiles for different groups of CMs. Generally, a group of CMs that have similar SNR performance will be
grouped into the same profile. When packets are encoded into FEC codewords and transmitted into the OFDM
spectrum, a path in the downstream for that packet is created so the PHY layer uses a profile to create a path at the
MAC layer.
There can be multiple paths from the CMTS to the same CM. Each path has a different profile. Profiles are typically
given a letter designation such as Profile A. Profile A is the boot profile that CMs first begin receiving when they
initialize and register. Either the forwarding engine or the DOCSIS QoS engine keeps a lookup table of exactly to
what path and what profile each packet needs to be assigned. This profile assignment is used to pick a convergence
layer buffer.
There is one convergence layer buffer per profile. These are shallow buffers that hold only a few packets so as to not
build up any significant latency. The output of these buffers is fed to the codeword builder. The codeword builder is
responsible for mapping DOCSIS frames into codewords. It is also responsible for balancing out the traffic flow
between all the profiles so that the latency budgets are observed.
The codeword builder uses the same profile for an entire codeword. It can change profiles at each codeword
boundary. The convergence layer buffers do not have to be serviced in any particular order. The DOCSIS MAC
layer has already rate-shaped the packet flow to the size of the OFDM channel, so all packets will fit. It is up to the
codeword builder to schedule the packets into the FEC codewords as it deems appropriate. Although rate shaping or
packet drops are not intended to be performed at the convergence layer, some queues could be treated as low latency
while other queues could be treated as high latency.
172
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Since the codeword builder is multiplexing at the codeword level, the packets in the convergence layer buffers are
naturally split across codeword boundaries and multiplexed together. The convergence layer buffers are packets in
bytes out. The codeword builder combines bytes from one buffer, adds FEC, and then using the profile modulation
vector, it maps the codeword onto one or more OFDM symbols (or partial symbols).
8.3.2
MAC Frame to Codewords
The downstream LDPC codeword shown in Figure 86 is referred to as (16200, 14400). This means that a full
codeword is 2025 bytes (16200 bits) that are divided into 225 bytes (1800 bits) of parity and 1800 bytes (14400 bits)
of LDPC payload. That payload is further divided into 21 bytes (168 bits) of BCH parity, a 2 byte fixed header, and
a variable 1777 byte maximum payload for DOCSIS frames. When the FEC codeword is shortened, only the
DOCSIS payload shrinks. All other fields remain the same size.
Figure 86 - DOCSIS Frame to Codeword Mapping
DOCSIS frames are sequentially mapped to codewords that are associated with a common profile. The codewords
do not need to be adjacent, although their order is guaranteed. If there is no DOCSIS frame available, the CMTS
MUST use a stuffing pattern of 0xFF. The codeword header is defined in Table 82.
Table 82 - Data Codeword Definition
Name
Length
Value
Valid
1 bit
0 = PDU Pointer is not valid (ignore)

1 = PDU pointer is valid
Reserved
4 bits
Set to 0. Ignore on receive.
PDU Pointer
11 bits
This points to the first byte of the first DOCSIS frame that starts in the payload. A value of
zero points to the byte immediately following the codeword header.
When the codeword gets mapped across subcarriers within a symbol, there may be residual bits left over on the last
subcarrier within that symbol. Since the number of residual bits may be more or less than 8, the receiver cannot
simply round down to a byte boundary. To permit the downstream receiver to discard these residual bits properly,
the CMTS MUST make the codeword payload an odd number of bytes.
One potential set of algorithms for the CMTS codeword builder and the CM codeword receiver is as follows.
On the CMTS, the algorithm is:
Number of total bytes is (header + payload + parity) and never exceeds 2025 bytes.
IF (total bytes = odd), send to FEC engine.
IF (total bytes = even), add a0xFF stuff byte to the payload if legal or change the number of bytes.
CMTS Symbol mapper adds trailing bits to map codeword to a symbol boundary.
10/29/13
CableLabs
173
CM-SP-PHYv3.1-I01-131029
On the CM, the corresponding algorithm is:
CM extracts total bits between two NCP pointers.
IF total bits > 16200, use initial 16200 bits, and a full codeword is declared.
IF total bits = 16200, and a full codeword is declared.
IF total bits < 16200, round down to the nearest odd number of bytes.
Discard [(total bits + 8) Modulo 16] bits.
8.3.3
Subcarrier Numbering Conventions
Subcarriers are numbered from lower frequency to higher frequency within a FFT block. All subcarriers within the
204.8 MHz bandwidth are numbered, including the outside excluded subcarriers. Numbering starts at 0 and goes to
4095 for 4K FFT, and 0 to 8191 for 8K FFT.
Data codewords are mapped to subcarriers - prior to time and frequency interleaving - from a lower number to a
higher number.
8.3.4
Next Codeword Pointer
When the data codewords are mapped to subcarriers within a symbol, a pointer is needed to identify where a data
codeword starts. This is known as the Next Codeword Pointer (NCP). There are a variable number of NCP message
blocks (MBs) on each OFDM symbol. To make sure that all subcarriers are used without reserving empty NCP
MBs, the mapping of the NCP occurs in the opposite direction of the mapping for data. The relationship of NCP
message blocks to the data channel is shown in Figure 87 (scattered pilots are not shown; last NCP MB is a CRC).
Figure 87 - Data and NCP Prior to Interleaving
The CMTS MUST map data subcarriers within a symbol, starting from a lower frequency and proceeding to a
higher frequency. The CMTS MUST map NCP message blocks starting at a higher frequency and moving to a lower
frequency.
174
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Figure 88 - NCP Message Block
The format of the NCP is illustrated in Figure 88 and defined in Table 83. Note that each three byte NCP MB is
mapped into a unique FEC codeword that has a 3 byte payload with 3 bytes of FEC. The last FEC codeword is then
followed by a 3 byte CRC-24-D (refer to Annex E) that is also placed in its own FEC block.
Table 83 - NCP Parameters
Field
Size
Description
Profile ID 4 bits
Profile ID for the data channel

0 = Profile A
1 = Profile B
15 = Profile P
1 bit
Zero Bit Loading

0 = subcarriers follow profile
1 = subcarriers are all zero-bit-loaded
1 bit
Data Profile Update

0 = use even profile
1 = use odd profile
1 bit
NCP Update
0 = use even profile
1 = use odd profile
This bit is equal to the LSB of the NCP profile change count. This bit refers to the
NCP profile usage for the next symbol rather than the current symbol.
1 bit
Last NCP Block

0 = This NCP is followed by another NCP.
1 = This is the last NCP in the chain and is followed by a CRC.
1 bit
Directed Test
0 = this codeword is not suitable for directed profile testing by CMs
1 = this codeword is suitable for directed profile testing by CMs
1 bit
Reserved
Subcarrier 13 bits
pointer
10/29/13
This is the number assigned to the first subcarrier used by the codeword. The
maximum value is 0x1FFE = 8190. The value 0x1FFF is reserved as a null pointer.
CableLabs
175
CM-SP-PHYv3.1-I01-131029
The NCP structure is predicated upon the following facts:
FEC codewords are mapped continuously across successive symbols.
The PHY can determine the first subcarrier of the first NCP message block
The PHY can determine the first subcarrier of the data field in the current symbol.
Based upon these facts and combined with the information in the NCP fields, then
The PHY can determine the last subcarrier of the last NCP message block
The next subcarrier after the last NCP message block CRC is last subcarrier of the data field.
The main task of the NCP message block is to provide a reference to the appropriate profile and a start pointer for
codewords. The length of a codeword is determined by the difference between the subcarrier pointer in two
successive NCP message blocks.
Data subcarriers may contain FEC codewords or unused sub-carriers. These functions are referred to as fields in the
NCP header.
The CMTS MUST include one NCP within the same symbol for each start of codeword or a group of unused
subcarriers that exists in that symbol. The CMTS MUST include a valid Profile ID when the field is a FEC
codeword field and no zero-bit-loading (Z bit not asserted).
The CMTS MUST assert the Zero load bit ("Z") to mark a set of subcarriers which are not used as described in
Section 8.3.4.1.
The CM MUST ignore the Profile ID when the Z bit is asserted.
The CM MUST use the Profile Update bit ("C") to select the odd or even data profile.
The CM MUST use the NCP Update bit ("N") to select the odd or even NCP profile. Since the profile for NCP has
to be known prior to receiving the NCP, this bit refers to the profile used for NCP in the next symbol. The CM
MUST use the NCP Update bit ("N") to select the odd or even NCP profile for the next OFDM symbol. The CMTS
MUST assert this bit to the same value in all the NCPs of any symbol.
The CMTS MUST assert the Last bit ("L") if the NCP is the last NCP message block. The CMTS MUST follow the
NCP block that has its Last bit asserted with a CRC-24-D. The CRC-24-D is calculated across all message blocks in
a symbol exclusive of the FEC parity bits. The CMTS MUST assert the Test indicator ("T") when the subcarriers
pointed to contain data that the CM should perform a test on.
A NULL NCP is defined as an NCP with the start pointer set to 0x1FFF. The usage of Null NCPs is defined in the
next section. An Active NCP is an NCP that points to valid FEC codeword. Therefore, an Active NCP is an NCP in
which Z-bit is Zero and the Start Pointer is not equal to 0x1FFF.
8.3.4.1
NCP Usage
The CMTS MUST NOT place more than 11 NCPs plus a CRC for a total of 12 NCP MBs in an 8K OFDM symbol.
The CMTS MUST NOT place more than 12 NCPs plus two CRCs for a total of 14 NCP MBs in any two successive
4K OFDM symbols.
In the case of an 8K FFT OFDM symbol, the 12 NCP MBs will be formed by a maximum of 10 active NCP MBs,
the NULL or zero-bit-loaded NCP MB (i.e. NCP with Z-bit set to ONE) and the NCP CRC MB.
In the case of a 4K FFT, in addition to the 10 maximum active NCP MBs over two successive symbols, each of
these symbols may have one additional NULL or zero-bit-loaded NCP MB, and each symbol will have a NCP CRC
MB. This brings the maximum number of NCP MBs over two successive symbols to 14.
If the data FEC blocks are small in one 8K FFT OFDM symbol, there could be data subcarriers left in the symbol
after the placement of 10 active NCPs and the corresponding data. In such a case the CMTS MUST include an NCP
describing the remaining subcarriers as zero-bit-loaded ("Z" bit asserted).
The CMTS MUST NOT place more than 10 active NCPs in any two consecutive 4K OFDM symbols, i.e., the
number of active NCPs in 4K FFT OFDM symbols n and n+1 MUST NOT exceed 10, for any value of n.
176
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
If the data FEC blocks are small, there could be data subcarriers remaining and unused after the placement of 10
active NCPs and the corresponding data in two consecutive 4K OFDM symbols. In such a case, the CMTS MUST
include an NCP describing the remaining subcarriers as zero-bit-loaded ("Z" bit asserted). Furthermore, in the case
of 4K FFT, if all of the 10 active NCPs are consumed by the symbol n, then the CMTS MUST place an NCP in
symbol n+1 indicating that the unused subcarriers are zero-bit-loaded. The symbol n+1 may contain a continuation
of a codeword from symbol n, but no new codeword can start in symbol n+1.
For small bandwidths it is possible that there may not be a beginning or an end of a FEC codeword in a symbol.
That is, a codeword may begin in the previous symbol and end in the following symbol. In such a case the CMTS
MUST insert a NULL NCP in the current symbol.
There may also be scenarios in which a FEC codeword may end within a symbol without leaving sufficient space to
include an NCP. In this case, the CMTS MUST insert a NULL NCP and move some of the data subcarriers of the
FEC codeword to next OFDM symbol.
8.3.4.2
NCP Example
Figure 89 - NCP Examples
Figure 89 shows some examples of how the NCP field is used. This view is prior to interleaving. NCP blocks are
mapped to sub-carriers starting with the first non-excluded subcarrier at the top of the spectrum and then down in
frequency. After the last NCP MB is a CRC-24-D. Data is mapped to the first non-excluded subcarrier at the bottom
of the frequency range and then continuing upwards in frequency.
In symbol 1,
codeword A starts at the beginning of the symbol and has a start pointer. Codeword B starts
after codeword A and has a start pointer. The length of codeword A is the difference between
the codeword A start pointer and the codeword B start pointer.
In symbol 2,
codeword C starts at the beginning of the symbol and has a start pointer. The length of the
previous codeword B is derived from the difference between the codeword B start pointer
and the codeword C start pointer, taking into account where the last data subcarrier was in
symbol 1. Codeword D gets a start pointer.
10/29/13
CableLabs
177
CM-SP-PHYv3.1-I01-131029
In symbol 3,
codeword D continues from symbol 2 and finishes. Codeword A follows and is given a start
pointer. The length of codeword D is derived from the difference between the codeword C
start pointer and the codeword D start pointer, taking into account where the last data
subcarrier was in symbol 2.
In symbol 4,
codeword A continues. Since there is no start pointer required, but at least one NCP block is
required, an NCP block with a null pointer is included.
In symbol 5,
codeword A ends. Codeword B begins and ends. A single NCP block is created with a start
pointer to codeword B.
In symbol 6,
codeword C both starts and ends. A single NCP block is created with a start pointer to
codeword C.
In symbol 7,
codeword D starts and ends. There are no more data packets to send, so the remaining
subcarriers are unused. A NCP block is assigned for the codeword D start pointer. A second
NCP block is assigned to the start pointer of the unused subcarriers. This start pointer is used
to determine the length of codeword D.
In symbol 8,
codeword A begins and ends. Codeword B begins and tried to end with a few subcarriers
unused between the end of the data codeword and the end of the NCP field. Since no
subcarriers can be left unused, and since an NCP would not fit, an NCP with a null pointer
was inserted and some of the last few bytes of codeword B were forced into the next symbol.
There is an NCP message block for codeword A, codeword B, and the null NCP.
In symbol 9,
codeword C starts a few subcarriers into the symbol. There is one NCP block for
codeword C.
178
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
9 PROACTIVE NETWORK MAINTENANCE

9.1
Scope
This section defines the requirements supporting Proactive Network Maintenance (PNM). CMTS and cable modem
features and capabilities can be leveraged to enable measurement and reporting of network conditions such that
undesired impacts such as plant equipment and cable faults, interference from other systems and ingress can be
detected and measured. With this information cable network operations personnel can make modifications necessary
to improve conditions and monitor network trends to detect when network improvements are needed.
9.1 Scope
Describes the structure of the DOCSIS 3.1 PHY specification Proactive Network Maintenance section.
9.2 System Description
Identifies the components, test points, and management capabilities of the DOCSIS 3.1 Proactive Network
Maintenance system.
9.3 Downstream PNM Requirements
Defines CMTS and CM requirements for obtaining and buffering symbol samples, triggering collection of
upstream spectrum conditions information, providing wideband spectrum analysis, employing excluded
subcarriers as a spectral notch, providing equalizer coefficient values, providing QAM constellation points for
display, obtaining and reporting receiver MER measurements, obtaining and reporting forward error correction
statistics, and reporting signal histograms for the downstream channel.
9.4 Upstream PNM Requirements
Defines CMTS and CM requirements for obtaining and buffering symbol samples, providing wideband
spectrum analysis, obtaining and reporting noise power measurements and statistics, providing equalizer
coefficient values, obtaining and reporting forward error correction statistics, and reporting signal histograms
for the upstream channel.
9.2
System Description
As shown in Figure 91, the CMTS and CM contain test points which include essential functions of a spectrum
analyzer, vector signal analyzer (VSA), and network analyzer, while the cable plant is considered the Device Under
Test (DUT). The goal is to rapidly and accurately characterize, maintain and troubleshoot the upstream and
downstream cable plant, in order to guarantee the highest throughput and reliability of service.
10/29/13
CableLabs
179
CM-SP-PHYv3.1-I01-131029
Figure 91 - Test points in CM and CMTS Supporting Proactive Network Maintenance
9.3
Downstream PNM Requirements
9.3.1
Downstream Symbol Capture
The purpose of downstream symbol capture is to provide partial functionality of a network analyzer to analyze the
response of the cable plant.
At the CMTS, the frequency-domain modulation values of one full OFDM symbol before the IFFT are captured and
made available for analysis. This includes the I and Q modulation values of all subcarriers, including data
subcarriers, pilots, PLC preamble symbols and excluded subcarriers. This capture will result in a number of data
points equal to the FFT length in use (4096 or 8192), 16 bits in width for each of I&Q, with LSBs padded with zeros
if required.
At the CM, the received I and Q time-domain samples of one full OFDM symbol before the FFT, not including the
guard interval, at the 204.8 MHz FFT sample rate, are captured and made available for analysis. This capture will
result in a number of data points equal to the FFT length in use (4096 or 8192), 16 bits in width for each of I&Q,
with LSBs padded with zeros if required. The capture will include a bit indicating if receiver windowing effects are
present in the data.
Capturing the input and output of the cable plant is equivalent to a wideband sweep of the channel, which permits
full characterization of the linear and nonlinear response of the downstream plant. The MAC provides signaling via
the PLC Trigger Message to ensure that the same symbol is captured at the CMTS and CM.
The CMTS MUST be capable of capturing the modulation values of one full downstream symbol for analysis.
The CM MUST be capable of locating and capturing the time-domain samples of one full downstream symbol for
analysis.
9.3.2
Downstream Wideband Spectrum Capture
The purpose of downstream wideband spectrum capture is to provide a downstream wideband spectrum analyzer
function in the DOCSIS 3.1 CM similar to the capability provided in DOCSIS 3.0.
180
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
The CM MUST provide a downstream wideband spectrum capture and analysis capability.
The CM SHOULD provide the capability to capture and analyze the full downstream band of the cable plant.
9.3.3
Downstream Noise Power Ratio (NPR) Measurement
The purpose of downstream NPR measurement is to view the noise, interference and intermodulation products
underlying a portion of the OFDM signal. The CMTS defines an exclusion band of zero-valued subcarriers which
forms a spectral notch in the downstream OFDM signal. The CM provides its normal spectral capture measurements
per 9.3.2, which show the notch depth. The maximum notch width is not specified, but will normally not exceed 10
MHz. A possible use case is to observe LTE interference occurring within an OFDM band; another is to observe
intermodulation products resulting from signal-level alignment issues.
The CMTS MUST be capable of accepting start and stop subcarrier indices defining an exclusion band (notch).
The CMTS MUST set the modulation value of all subcarriers in the notch to zero (no energy).
9.3.4
Downstream Equalizer Coefficients
The purpose of equalizer coefficients is to provide access to CM downstream adaptive equalizer coefficients, which
describe the linear response of the cable plant. The OSSI spec will define summary metrics to avoid having to send
all equalizer coefficients on every query.
The CM MUST report its downstream adaptive equalizer coefficients (full set or summary) for any single OFDM
block upon request.
9.3.5
Downstream Constellation Display
The downstream constellation display provides received QAM constellation points for display. Equalized soft
decisions (I and Q) at the slicer input are collected over time, possibly subsampling to reduce complexity, and made
available for analysis. Start and stop indices define the range of subcarriers which are included in the measurement.
Only data-bearing subcarriers with the specified profile and QAM constellation are sampled. Pilots and excluded
subcarriers within the range are ignored. At least 8192 samples are provided for each query; additional queries may
be made to further fill in the plot.
The CM MUST be capable of capturing and reporting received soft-decision samples, for a single selected profile,
single constellation, and selectable range of subcarriers within a single OFDM Block.
9.3.6
Downstream Receive Modulation Error Ratio (RxMER) Per Subcarrier
The downstream receive modulation error ratio (RxMER) per subcarrier provides measurements of the receive
modulation error ratio (RxMER) for each subcarrier. The CM measures the RxMER using pilots and zero-valued
subcarriers, which are not subject to symbol errors as data subcarriers would be. Since scattered pilots visit all data
subcarriers, and zero-valued subcarriers are located in defined locations including exclusion bands, the RxMER of
all subcarriers in the active OFDM band can be measured over time. The scattered pilot pattern overlaps the PLC
preamble symbols, which are used for the measurement as if they were pilots. Only those zero-valued subcarriers
which are processed by the CM receiver are measured. For the purposes of this measurement, RxMER is defined as
the ratio of the average power of the equalized QAM constellation to the average error-vector power. For pilots, the
error vector is the difference between the equalized received pilot value and the known correct pilot value. For zerovalued subcarriers, the error vector is the unequalized received value itself, since the correct value is zero and there
is no reliable channel estimate for excluded subcarrier locations with which to perform equalization. Using this
definition, the noise measurement of a zero-valued subcarrier is expressed in terms of an equivalent RxMER value
using the average QAM constellation power as a reference. As a defining test case, for an ideal AWGN channel, an
OFDM block containing a mix of QAM constellations, including some zero-valued subcarriers, with 35 dB CNR on
the QAM subcarriers, will yield an RxMER measurement of nominally 35 dB for all subcarrier locations including
the zero-valued subcarriers.
The CM MUST be capable of providing measurements of the RxMER for all subcarrier locations for a single
OFDM Block, using pilots, PLC preamble symbols, and/or zero-valued subcarriers for the measurement. The CM
MAY omit measurements on some zero-valued subcarriers.
10/29/13
CableLabs
181
CM-SP-PHYv3.1-I01-131029
9.3.6.1
Signal-to-Noise Ratio (SNR) Margin for Candidate Profile
The purpose of this item is to provide an estimate of the SNR margin available on the downstream data channel with
respect to a candidate modulation profile. The following algorithm is used to compute this estimate. The CM only
performs this computation upon request. The same computation is done for the NCP channel.
Algorithm: The CM measures the RxMER value for each data subcarrier as specified in Section 9.3.6. From these
measurements it calculates the average RxMER per data subcarrier, MER1. It accepts as an input the required
average MER per subcarrier for the candidate profile, MER2. The SNR margin is defined as MER1 MER2, where
all quantities are in dB. As an example, if the CM measures MER1 = 33 dB, and the candidate profile requires
MER2 = 30 dB, the CM reports an SNR margin of 3 dB. In addition, the CM reports the number of subcarriers
whose RxMER is at least x dB below the threshold of CER = 1e-5 for a given QAM order, where x is a configurable
parameter with default value = 3.
9.3.7
Downstream FEC Statistics
The purpose of the FEC statistics is to monitor downstream link quality via FEC and related statistics. Statistics are
taken on FEC codeword error events, taking into account both the inner LDPC code and outer BCH code, and are
provided on each OFDM channel and for each profile being received by the CM. The measurements are
timestamped using bits 21-52 of the 64-bit extended timestamp, where bit 0 is the LSB, which provides a 32-bit
timestamp value with resolution of 0.4 ms and range of 20 days. Timestamping is done with nominal accuracy of
100 ms or better. Codeword counts and codeword error counts include only full-length codewords, i.e., having
LDPC codewords of size 16,200 bits. Similar statistics are taken on the NCP, also only using full-length codewords,
and on the PLC. MAC packet statistics are not profile-based, but are computed on all packets addressed to the CM.
The CM MUST be capable of providing the following downstream performance metrics:
Uncorrectable full-length codewords: Number of full-length codewords that failed BCH decoding.
Correctable full-length codewords: Number of full-length codewords that failed pre-decoding LDPC
syndrome check and passed BCH decoding.
Unreliable full-length NCP Codewords: Number of full-length NCP codewords that failed LDPC postdecoding syndrome check.
Unreliable PLC Codewords: Number of PLC codewords that failed LDPC post-decoding syndrome
check.
NCP full-length CRC failures: Number of full-length NCP codewords that failed CRC check.
MAC CRC failures: Number of packets that failed MAC CRC check.
Total number of full-length FEC codewords.
Total number of full-length NCP codewords.
Total number of PLC codewords.
Total number of MAC packets.
Start and stop time of analysis period.
The CM MUST be capable of providing the following downstream FEC summaries on each OFDM channel for
each profile being received by the CM:
182
Full-length codeword error error ratio vs. time (seconds): Ratio of full-length number of uncorrectable
codewords to total number of full-length codewords in each one-second interval for a rolling 10-minute
period (600 values).
full-length codeword error ratio vs. time (minutes): Ratio of number of uncorrectable full-length
codewords to total number of full-length codewords in each one-minute interval for a rolling 24-hour
period (1440 values).
Ending time of rolling period.
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Red/yellow/green summary link status (colors defined in [DOCSIS OSSIv3.1]).
The CM MUST provide two collection and reporting methods for each error-count metric:
9.3.8
Long-term statistics. The CM always collects metrics in the background for each profile. The codeword
(or packet) and error counters are automatically reset once per hour. The previous value of each counter
is saved when the counter is reset due to timeout, so that in a steady-state condition a full hours reading
is always available.
Short-term statistics. The CM performs a one-shot measurement with two configured parameters, Ne and
Nc. The CM reports the results when Ne errors have occurred or Nc codewords have been processed,
whichever comes first. This measurement is particularly useful for downstream profile performance
measurement.
Downstream Histogram
The purpose of downstream histograms is to provide a measurement of nonlinear effects in the channel such as
amplifier compression and laser clipping. For example, laser clipping causes one tail of the histogram to be
truncated and replaced with a spike. The CM captures the histogram of time domain samples at the wideband front
end of the receiver. The histogram is two-sided; that is, it encompasses values from far-negative to far-positive
values of the samples.
The CM MUST be capable of capturing the histogram of time domain samples at the wideband front end of the
receiver. The histogram MUST have a programmable accumulation period of 1 second to 1 minute and a minimum
resolution of 255 bins.
9.3.9
Downstream Profile Performance Metrics
The following measurements are used for both PNM diagnostics and for performance testing of downstream
modulation profiles. The data format used to report a given measurement may differ for these two uses.
Uncorrectable full-length codewords
Corrected full-length codewords
MAC CRC codeword errors
NCP LDPC full-length codeword errors
NCP full-length CRC failures
Total number of full-length FEC codewords
Total number of full-length NCP codewords
RxMER per subcarrier
RxMER measurement type per subcarrier
SNR margin for candidate data profile
SNR margin for candidate NCP profile
9.4
Upstream PNM Requirements
9.4.1
Upstream Capture for Quiet Period and Probe Symbol
The purpose of the capture is to view underlying noise and measure plant response, by capturing one OFDM symbol
during a scheduled quiet period or probe. A quiet period provides an opportunity to view the underlying noise and
ingress while no traffic is being transmitted in the OFDM band. An upstream probe provides the partial functionality
of a network analyzer, since the input is known and the output is captured. This permits full characterization of the
linear and nonlinear response of the upstream cable plant. A list of excluded subcarriers is also provided, in order to
10/29/13
CableLabs
183
CM-SP-PHYv3.1-I01-131029
fully define the transmitted waveform. The index of the starting sample used by the receiver for its FFT is also
reported.
The CMTS MUST be capable of capturing the samples of one upstream OFDM symbol, including guard time,
during an upstream quiet period or probe, and making them available for analysis.
9.4.2
Upstream Triggered Spectrum Capture
The upstream triggered spectrum capture provides a wideband spectrum analyzer function in the CMTS which can
be triggered to examine desired upstream transmissions as well as underlying noise/interference during a quiet
period. The capture capability herein covers both OFDM and pre-DOCSIS-3.1 upstream channels that may exist in
the Upstream Spectrum.
The CMTS MUST provide wideband spectrum analysis capability.
The CMTS SHOULD provide a spectrum analysis span covering up to the full upstream spectrum of the cable plant.
The CMTS MUST provide the ability to trigger the spectrum sample capture using the following modes:
Free running
Trigger on SID (service identifier)
Trigger during quiet period
Trigger on minislot count.
9.4.3
Upstream Impulse Noise Statistics
Upstream Impulse noise statistics gather statistics of burst/impulse noise occurring in a selected narrow band. A
bandpass filter is positioned in an unoccupied upstream band. A threshold is set, energy exceeding the threshold
triggers the measurement of an event, and energy falling below the threshold ends the event. The threshold may be
set to zero, in which case the average power in the band will be measured. The measurement is time-stamped using
the D3.0 field of the 64-bit extended timestamp (bits 9-40, where bit 0 is the LSB), which provides a resolution of
98 ns and a range of 7 minutes.
The CMTS MUST provide the capability to capture the following statistics in a selected band up to 5.12 MHz wide:
Timestamp of event
Duration of event
Average power of event.
The CMTS MUST provide a time history buffer of up to 1024 events.

9.4.4
Upstream Equalizer Coefficients
Upstream equalizer coefficients provide access to CM upstream pre-equalizer coefficients, and CMTS upstream
adaptive equalizer (post-equalizer) coefficients, which taken together describe the linear response of the upstream
cable plant for a given CM. The OSSI spec will define summary metrics to avoid having to send all equalizer
coefficients on every query.
The CM MUST provide the capability to report its upstream pre-equalizer coefficients (full set or summary) upon
request.
The CMTS MUST provide a capability for reporting its upstream adaptive equalizer coefficients associated with a
given CM upon request.
9.4.5
Upstream FEC Statistics
Upstream FEC statistics provide for monitoring upstream link quality via FEC and related statistics. Statistics are
taken on codeword error events. The measurement is time-stamped using bits 21-52 of the extended timestamp. An
LDPC codeword that fails post-decoding syndrome check will be labeled "unreliable", but the data portion of the
184
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
codeword may not contain bit errors; hence the "unreliable codeword" count will tend to be pessimistic. All
codewords, whether full-length or shortened, are included in the measurements.
The CMTS MUST be capable of providing the following FEC statistics for any single upstream user:
Pre-FEC Error-Free Codewords: Number of codewords that passed pre-decoding syndrome check.
Unreliable Codewords: Number of codewords that failed post-decoding syndrome check.
Corrected Codewords: Number of codewords that failed pre-decoding syndrome check, but passed postdecoding syndrome check.
MAC CRC failures: Number of packets that failed MAC CRC check.
Total number of FEC codewords.
Total number of MAC packets.
Start and stop time of analysis period.
The CMTS MUST be capable of providing the following FEC summaries over a period of up to 10 minutes for any
single upstream user:
9.4.6
Total number of seconds
Number of errored seconds (seconds during which at least one unreliable codeword occurred)
Count of codeword errors (unreliable codewords) in each 1-second interval
Start and stop time of summary period
Red/yellow/green summary link status (colors defined in [DOCSIS OSSIv3.1]).

Upstream Histogram
The purpose of upstream histograms is to provide a measurement of nonlinear effects in the upstream channel such
as amplifier compression and laser clipping. For example, laser clipping causes one tail of the histogram to be
truncated and replaced with a spike. The CMTS captures the histogram of time domain samples at the wideband
front end of the receiver. The histogram is two-sided; that is, it encompasses values from far-negative to far-positive
values of the samples.
The CMTS MUST be capable of capturing the histogram of time domain samples at the wideband front end of the
receiver. The histogram MUST have a programmable accumulation period of 1 second to 1 minute and a minimum
resolution of 255 bins.
10/29/13
CableLabs
185
CM-SP-PHYv3.1-I01-131029
Annex A
QAM Constellation Mappings
The CMTS MUST use the QAM constellation mappings given in this section for all downstream transmissions.
Downstream transmissions do not contain BPSK, 8-QAM and 32-QAM constellations.
The CM MUST use the QAM constellation mappings given in this section for all upstream transmissions. Upstream
transmissions do not contain 8192-QAM and 16384-QAM constellations.
A.1
QAM Constellations
The figures given below show the mapping of an m-tuple onto a (Real, Imaginary) point in the complex plane. The
horizontal axis is the real axis and the vertical axis is the imaginary axis.
The m-tuple is represented by:
{0 1 1 }
Mapping of the FEC encoded bit streams to the m-tuples is described in the sections detailing downstream and
upstream transmissions.
Each m-tuple is represented as a hexadecimal number in all the constellation diagrams given below.
-1
Figure A1 - BPSK Constellation Mapping of { }
-1
-1
Figure A2 - QPSK Constellation Mapping of { }
186
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
3
5
3
1
-1
4
-3
2
0
-3
-1
Figure A3 - 8-QAM Constellation Mapping of { }
-1
-3
-1
-3
10
12
02
00
11
15
14
04
05
01
13
17
16
06
07
03
-1
1B
1F
1E
0E
0F
0B
-3
19
1D
1C
0C
0D
09
18
1A
0A
08
-5
-5
-3
-1
10/29/13
CableLabs
187
CM-SP-PHYv3.1-I01-131029
20
22
2A
28
08
0A
02
00
21
23
2B
29
09
0B
03
01
25
27
2F
2D
0D
0F
07
05
24
26
2E
2C
0C
0E
06
04
-1
34
36
3E
3C
1C
1E
16
14
-3
35
37
3F
3D
1D
1F
17
15
-5
31
33
3B
39
19
1B
13
11
-7
30
32
3A
38
18
1A
12
10
-1
-3
-5
-7
11
42
43
4B
4A
0A
0B
03
02
40
41
49
48
08
09
01
00
45
44
54
55
51
50
10
11
15
14
04
05
47
46
56
57
53
52
12
13
17
16
06
07
4F
4E
5E
5F
5B
5A
1A
1B
1F
1E
0E
0F
4D
4C
5C
5D
59
58
18
19
1D
1C
0C
0D
-1
6D
6C
7C
7D
79
78
38
39
3D
3C
2C
2D
-3
6F
6E
7E
7F
7B
7A
3A
3B
3F
3E
2E
2F
-5
67
66
76
77
73
72
32
33
37
36
26
27
-7
65
64
74
75
71
70
30
31
35
34
24
25
-9
60
61
69
68
28
29
21
20
-11
62
63
6B
6A
2A
2B
23
22
-11
-9
-7
-5
-3
-1
11
Figure A7 - 128-QAM Constellation mapping of { }
In order to reduce the size of the diagrams, only the first quadrant is shown for the larger constellations, namely,
256-QAM, 512-QAM, 1024-QAM, 2048-QAM, 4096-QAM, 8192-QAM and 16384-QAM. The mapping of the two
bits {0 1 } of {0 1 1 } is the same for all these QAM constellations, as illustrated in Figure A8.
188
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
10
00
11
01
Figure A8 - Mapping of Bits { } of { , , } for Constellations

with only one Quadrant Defined
The mapping of the bits of {2 3 , , 1 } to the first quadrant of the constellation is given in the figures below for
256-QAM, 512-QAM, 1024-QAM, 2048-QAM, 4096-QAM, 8192-QAM and 16384-QAM. The mappings for the
other three quadrants are obtained by mirroring the first quadrant about the horizontal and vertical axis as illustrated
in the figure below. This figure shows only a (3x3) grid of points in each quadrant for illustration of the above
mentioned reflective property. However, this reflective mapping is applicable to any number of points in each
quadrant. Quadrant 1 is reflected about the vertical axis to get quadrant 2. Quadrant 1 is reflected about the
horizontal axis to get quadrant 4. Quadrant 2 is reflected about the horizontal axis to get quadrant 3.
Quadrant 2
Quadrant 1
a(2,2) a(1,2) a(0,2)
a(0,2) a(1,2) a(2,2)
a(2,1) a(1,1) a(0,1)
a(0,1) a(1,1) a(2,1)
a(2,0) a(1,0) a(0,0)
a(0,0) a(1,0) a(2,0)
a(2,0) a(1,0) a(0,0)
a(0,0) a(1,0) a(2,0)
a(2,1) a(1,1) a(0,1)
a(0,1) a(1,1) a(2,1)
a(2,2) a(1,2) a(0,2)
a(0,2) a(1,2) a(2,2)
Quadrant 3
Quadrant 4
Figure A9 - Reflective Mapping of bits { , , } for All Constellations (except BPSK)
10/29/13
CableLabs
189
CM-SP-PHYv3.1-I01-131029
15
20
22
2A
28
08
0A
02
00
13
21
23
2B
29
09
0B
03
01
11
25
27
2F
2D
0D
0F
07
05
24
26
2E
2C
0C
0E
06
04
34
36
3E
3C
1C
1E
16
14
35
37
3F
3D
1D
1F
17
15
31
33
3B
39
19
1B
13
11
30
32
3A
38
18
1A
12
10
11
13
15
Figure A10 - 256-QAM Constellation Mapping of { } on to Quadrant 1
23
028
029
02D 02C 00C 00D 009
008
21
02A 02B 02F 02E 00E 00F 00B 00A
19
022
023
027
026
006
007
003
002
17
020
021
025
024
004
005
001
000
15
040
041
045
044
054
055
051
050
010
011
015
014
13
042
043
047
046
056
057
053
052
012
013
017
016
11
04A 04B 04F 04E 05E 05F 05B 05A 01A 01B 01F 01E
048
049
04D 04C 05C 05D 059
058
018
019
01D 01C
068
069
06D 06C 07C 07D 079
078
038
039
03D 03C
06A 06B 06F 06E 07E 07F 07B 07A 03A 03B 03F 03E
062
063
067
066
076
077
073
072
032
033
037
036
060
061
065
064
074
075
071
070
030
031
035
034
11
13
15
17
19
21
23
190
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
31
080
082 08A 088
0A8 0AA 0A2 0A0 020
022 02A 028
008
00A 002
000
29
081
083 08B 089
0A9 0AB 0A3 0A1 021
023 02B 029
009
00B 003
001
27
085
087 08F
02D 00D 00F 007
005
25
084
086 08E 08C 0AC 0AE 0A6 0A4 024
026 02E 02C 00C 00E 006
004
23
094
096 09E 09C 0BC 0BE 0B6 0B4 034
036 03E 03C 01C 01E 016
014
21
095
097 09F
037 03F
015
19
091
093 09B 099
0B9 0BB 0B3 0B1 031
033 03B 039
019
01B 013
011
17
090
092 09A 098
0B8 0BA 0B2 0B0 030
032 03A 038
018
01A 012
010
15
0D0 0D2 0DA 0D8 0F8 0FA 0F2 0F0 070
072 07A 078
058
05A 052
050
13
0D1 0D3 0DB 0D9 0F9 0FB 0F3 0F1 071
073 07B 079
059
05B 053
051
11
0D5 0D7 0DF 0DD 0FD 0FF 0F7 0F5 075
077 07F
07D 05D 05F 057
055
0D4 0D6 0DE 0DC 0FC 0FE 0F6 0F4 074
076 07E 07C 05C 05E 056
054
0C4 0C6 0CE 0CC 0EC 0EE 0E6 0E4 064
066 06E 06C 04C 04E 046
044
0C5 0C7 0CF 0CD 0ED 0EF 0E7 0E5 065
067 06F
045
0C1 0C3 0CB 0C9 0E9 0EB 0E3 0E1 061
063 06B 069
049
04B 043
041
0C0 0C2 0CA 0C8 0E8 0EA 0E2 0E0 060
062 06A 068
048
04A 042
040
08D 0AD 0AF 0A7 0A5 025
09D 0BD 0BF 0B7 0B5 035
11
13
15
027 02F
19
17
21
03D 01D 01F 017
06D 04D 04F 047
23
25
27
29
31
10/29/13
CableLabs
191
CM-SP-PHYv3.1-I01-131029
47
0A0 0A1 0A5 0A4 0B4 0B5 0B1 0B0 030 031 035 034 024 025 021 020
45
43
41
0A2 0A3 0A7 0A6 0B6 0B7 0B3 0B2 032 033 037 036 026 027 023 022
39
37
35
33
31
29
27
25
23
21
19
17
15
13
11
9
7
5
088 089 08D 08C 09C 09D 099 098 018 019 01D 01C 00C 00D 009 008
3
1
0AA 0AB 0AF 0AE 0BE 0BF 0BB 0BA 03A 03B 03F 03E 02E 02F 02B 02A
0A8 0A9 0AD 0AC 0BC 0BD 0B9 0B8 038 039 03D 03C 02C 02D 029 028
08A 08B 08F 08E 09E 09F 09B 09A 01A 01B 01F 01E 00E 00F 00B 00A
082 083 087 086 096 097 093 092 012 013 017 016 006 007 003 002
080 081 085 084 094 095 091 090 010 011 015 014 004 005 001 000
100 101 105 104 114 115 111 110 150 151 155 154 144 145 141 140 040 041 045 044 054 055 051 050
102 103 107 106 116 117 113 112 152 153 157 156 146 147 143 142 042 043 047 046 056 057 053 052
10A 10B 10F 10E 11E 11F 11B 11A 15A 15B 15F 15E 14E 14F 14B 14A 04A 04B 04F 04E 05E 05F 05B 05A
108 109 10D 10C 11C 11D 119 118 158 159 15D 15C 14C 14D 149 148 048 049 04D 04C 05C 05D 059 058
128 129 12D 12C 13C 13D 139 138 178 179 17D 17C 16C 16D 169 168 068 069 06D 06C 07C 07D 079 078
12A 12B 12F 12E 13E 13F 13B 13A 17A 17B 17F 17E 16E 16F 16B 16A 06A 06B 06F 06E 07E 07F 07B 07A
122 123 127 126 136 137 133 132 172 173 177 176 166 167 163 162 062 063 067 066 076 077 073 072
120 121 125 124 134 135 131 130 170 171 175 174 164 165 161 160 060 061 065 064 074 075 071 070
1A0 1A1 1A5 1A4 1B4 1B5 1B1 1B0 1F0 1F1 1F5 1F4 1E4 1E5 1E1 1E0 0E0 0E1 0E5 0E4 0F4 0F5 0F1 0F0
1A2 1A3 1A7 1A6 1B6 1B7 1B3 1B2 1F2 1F3 1F7 1F6 1E6 1E7 1E3 1E2 0E2 0E3 0E7 0E6 0F6 0F7 0F3 0F2
1AA 1AB 1AF 1AE 1BE 1BF 1BB 1BA 1FA 1FB 1FF 1FE 1EE 1EF 1EB 1EA 0EA 0EB 0EF 0EE 0FE 0FF 0FB 0FA
1A8 1A9 1AD 1AC 1BC 1BD 1B9 1B8 1F8 1F9 1FD 1FC 1EC 1ED 1E9 1E8 0E8 0E9 0ED 0EC 0FC 0FD 0F9 0F8
188 189 18D 18C 19C 19D 199 198 1D8 1D9 1DD 1DC 1CC 1CD 1C9 1C8 0C8 0C9 0CD 0CC 0DC 0DD 0D9 0D8
18A 18B 18F 18E 19E 19F 19B 19A 1DA 1DB 1DF 1DE 1CE 1CF 1CB 1CA 0CA 0CB 0CF 0CE 0DE 0DF 0DB 0DA
182 183 187 186 196 197 193 192 1D2 1D3 1D7 1D6 1C6 1C7 1C3 1C2 0C2 0C3 0C7 0C6 0D6 0D7 0D3 0D2
180 181 185 184 194 195 191 190 1D0 1D1 1D5 1D4 1C4 1C5 1C1 1C0 0C0 0C1 0C5 0C4 0D4 0D5 0D1 0D0
11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47
192
CableLabs
10/29/13
63
61
59
57
55
53
51
49
47
45
43
41
39
37
35
33
31
29
27
25
23
21
19
17
15
13
11
9
7
5
3
1
CM-SP-PHYv3.1-I01-131029
200 202 20A 208 228 22A 222 220 2A0 2A2 2AA 2A8 288 28A 282 280 080 082 08A 088 0A8 0AA 0A2 0A0 020 022 02A 028 008 00A 002 000
201 203 20B 209 229 22B 223 221 2A1 2A3 2AB 2A9 289 28B 283 281 081 083 08B 089 0A9 0AB 0A3 0A1 021 023 02B 029 009 00B 003 001
205 207 20F 20D 22D 22F 227 225 2A5 2A7 2AF 2AD 28D 28F 287 285 085 087 08F 08D 0AD 0AF 0A7 0A5 025 027 02F 02D 00D 00F 007 005
204 206 20E 20C 22C 22E 226 224 2A4 2A6 2AE 2AC 28C 28E 286 284 084 086 08E 08C 0AC 0AE 0A6 0A4 024 026 02E 02C 00C 00E 006 004
214 216 21E 21C 23C 23E 236 234 2B4 2B6 2BE 2BC 29C 29E 296 294 094 096 09E 09C 0BC 0BE 0B6 0B4 034 036 03E 03C 01C 01E 016 014
215 217 21F 21D 23D 23F 237 235 2B5 2B7 2BF 2BD 29D 29F 297 295 095 097 09F 09D 0BD 0BF 0B7 0B5 035 037 03F 03D 01D 01F 017 015
211 213 21B 219 239 23B 233 231 2B1 2B3 2BB 2B9 299 29B 293 291 091 093 09B 099 0B9 0BB 0B3 0B1 031 033 03B 039 019 01B 013 011
210 212 21A 218 238 23A 232 230 2B0 2B2 2BA 2B8 298 29A 292 290 090 092 09A 098 0B8 0BA 0B2 0B0 030 032 03A 038 018 01A 012 010
250 252 25A 258 278 27A 272 270 2F0 2F2 2FA 2F8 2D8 2DA 2D2 2D0 0D0 0D2 0DA 0D8 0F8 0FA 0F2 0F0 070 072 07A 078 058 05A 052 050
251 253 25B 259 279 27B 273 271 2F1 2F3 2FB 2F9 2D9 2DB 2D3 2D1 0D1 0D3 0DB 0D9 0F9 0FB 0F3 0F1 071 073 07B 079 059 05B 053 051
255 257 25F 25D 27D 27F 277 275 2F5 2F7 2FF 2FD 2DD 2DF 2D7 2D5 0D5 0D7 0DF 0DD 0FD 0FF 0F7 0F5 075 077 07F 07D 05D 05F 057 055
254 256 25E 25C 27C 27E 276 274 2F4 2F6 2FE 2FC 2DC 2DE 2D6 2D4 0D4 0D6 0DE 0DC 0FC 0FE 0F6 0F4 074 076 07E 07C 05C 05E 056 054
244 246 24E 24C 26C 26E 266 264 2E4 2E6 2EE 2EC 2CC 2CE 2C6 2C4 0C4 0C6 0CE 0CC 0EC 0EE 0E6 0E4 064 066 06E 06C 04C 04E 046 044
245 247 24F 24D 26D 26F 267 265 2E5 2E7 2EF 2ED 2CD 2CF 2C7 2C5 0C5 0C7 0CF 0CD 0ED 0EF 0E7 0E5 065 067 06F 06D 04D 04F 047 045
241 243 24B 249 269 26B 263 261 2E1 2E3 2EB 2E9 2C9 2CB 2C3 2C1 0C1 0C3 0CB 0C9 0E9 0EB 0E3 0E1 061 063 06B 069 049 04B 043 041
240 242 24A 248 268 26A 262 260 2E0 2E2 2EA 2E8 2C8 2CA 2C2 2C0 0C0 0C2 0CA 0C8 0E8 0EA 0E2 0E0 060 062 06A 068 048 04A 042 040
340 342 34A 348 368 36A 362 360 3E0 3E2 3EA 3E8 3C8 3CA 3C2 3C0 1C0 1C2 1CA 1C8 1E8 1EA 1E2 1E0 160 162 16A 168 148 14A 142 140
341 343 34B 349 369 36B 363 361 3E1 3E3 3EB 3E9 3C9 3CB 3C3 3C1 1C1 1C3 1CB 1C9 1E9 1EB 1E3 1E1 161 163 16B 169 149 14B 143 141
345 347 34F 34D 36D 36F 367 365 3E5 3E7 3EF 3ED 3CD 3CF 3C7 3C5 1C5 1C7 1CF 1CD 1ED 1EF 1E7 1E5 165 167 16F 16D 14D 14F 147 145
344 346 34E 34C 36C 36E 366 364 3E4 3E6 3EE 3EC 3CC 3CE 3C6 3C4 1C4 1C6 1CE 1CC 1EC 1EE 1E6 1E4 164 166 16E 16C 14C 14E 146 144
354 356 35E 35C 37C 37E 376 374 3F4 3F6 3FE 3FC 3DC 3DE 3D6 3D4 1D4 1D6 1DE 1DC 1FC 1FE 1F6 1F4 174 176 17E 17C 15C 15E 156 154
355 357 35F 35D 37D 37F 377 375 3F5 3F7 3FF 3FD 3DD 3DF 3D7 3D5 1D5 1D7 1DF 1DD 1FD 1FF 1F7 1F5 175 177 17F 17D 15D 15F 157 155
351 353 35B 359 379 37B 373 371 3F1 3F3 3FB 3F9 3D9 3DB 3D3 3D1 1D1 1D3 1DB 1D9 1F9 1FB 1F3 1F1 171 173 17B 179 159 15B 153 151
350 352 35A 358 378 37A 372 370 3F0 3F2 3FA 3F8 3D8 3DA 3D2 3D0 1D0 1D2 1DA 1D8 1F8 1FA 1F2 1F0 170 172 17A 178 158 15A 152 150
310 312 31A 318 338 33A 332 330 3B0 3B2 3BA 3B8 398 39A 392 390 190 192 19A 198 1B8 1BA 1B2 1B0 130 132 13A 138 118 11A 112 110
311 313 31B 319 339 33B 333 331 3B1 3B3 3BB 3B9 399 39B 393 391 191 193 19B 199 1B9 1BB 1B3 1B1 131 133 13B 139 119 11B 113 111
315 317 31F 31D 33D 33F 337 335 3B5 3B7 3BF 3BD 39D 39F 397 395 195 197 19F 19D 1BD 1BF 1B7 1B5 135 137 13F 13D 11D 11F 117 115
314 316 31E 31C 33C 33E 336 334 3B4 3B6 3BE 3BC 39C 39E 396 394 194 196 19E 19C 1BC 1BE 1B6 1B4 134 136 13E 13C 11C 11E 116 114
304 306 30E 30C 32C 32E 326 324 3A4 3A6 3AE 3AC 38C 38E 386 384 184 186 18E 18C 1AC 1AE 1A6 1A4 124 126 12E 12C 10C 10E 106 104
305 307 30F 30D 32D 32F 327 325 3A5 3A7 3AF 3AD 38D 38F 387 385 185 187 18F 18D 1AD 1AF 1A7 1A5 125 127 12F 12D 10D 10F 107 105
301 303 30B 309 329 32B 323 321 3A1 3A3 3AB 3A9 389 38B 383 381 181 183 18B 189 1A9 1AB 1A3 1A1 121 123 12B 129 109 10B 103 101
300 302 30A 308 328 32A 322 320 3A0 3A2 3AA 3A8 388 38A 382 380 180 182 18A 188 1A8 1AA 1A2 1A0 120 122 12A 128 108 10A 102 100
9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63
10/29/13
CableLabs
193
CM-SP-PHYv3.1-I01-131029
95
93
91
89
87
85
83
81
79
77
75
73
71
69
67
65
63
61
59
57
55
53
51
49
47
45
43
41
39
37
35
33
31
29
27
25
23
21
19
17
15
13
11
9
7
5
3
1
0280
0282
028A
0288
02A8
02AA
02A2
02A0
0220
0222
022A
0228
0208
020A
0202
0200
0400
0402
040A
0408
0428
042A
0422
0420
04A0
04A2
04AA
04A8
0488
048A
0482
0480
0680
0682
068A
0688
06A8
06AA
06A2
06A0
0620
0622
062A
0628
0608
060A
0602
0600
0281
0283
028B
0289
02A9
02AB
02A3
02A1
0221
0223
022B
0229
0209
020B
0203
0201
0401
0403
040B
0409
0429
042B
0423
0421
04A1
04A3
04AB
04A9
0489
048B
0483
0481
0681
0683
068B
0689
06A9
06AB
06A3
06A1
0621
0623
062B
0629
0609
060B
0603
0601
0285
0287
028F
028D
02AD
02AF
02A7
02A5
0225
0227
022F
022D
020D
020F
0207
0205
0405
0407
040F
040D
042D
042F
0427
0425
04A5
04A7
04AF
04AD
048D
048F
0487
0485
0685
0687
068F
068D
06AD
06AF
06A7
06A5
0625
0627
062F
062D
060D
060F
0607
0605
0284
0286
028E
028C
02AC
02AE
02A6
02A4
0224
0226
022E
022C
020C
020E
0206
0204
0404
0406
040E
040C
042C
042E
0426
0424
04A4
04A6
04AE
04AC
048C
048E
0486
0484
0684
0686
068E
068C
06AC
06AE
06A6
06A4
0624
0626
062E
062C
060C
060E
0606
0604
0294
0296
029E
029C
02BC
02BE
02B6
02B4
0234
0236
023E
023C
021C
021E
0216
0214
0414
0416
041E
041C
043C
043E
0436
0434
04B4
04B6
04BE
04BC
049C
049E
0496
0494
0694
0696
069E
069C
06BC
06BE
06B6
06B4
0634
0636
063E
063C
061C
061E
0616
0614
0295
0297
029F
029D
02BD
02BF
02B7
02B5
0235
0237
023F
023D
021D
021F
0217
0215
0415
0417
041F
041D
043D
043F
0437
0435
04B5
04B7
04BF
04BD
049D
049F
0497
0495
0695
0697
069F
069D
06BD
06BF
06B7
06B5
0635
0637
063F
063D
061D
061F
0617
0615
11
0291
0293
029B
0299
02B9
02BB
02B3
02B1
0231
0233
023B
0239
0219
021B
0213
0211
0411
0413
041B
0419
0439
043B
0433
0431
04B1
04B3
04BB
04B9
0499
049B
0493
0491
0691
0693
069B
0699
06B9
06BB
06B3
06B1
0631
0633
063B
0639
0619
061B
0613
0611
0290
0292
029A
0298
02B8
02BA
02B2
02B0
0230
0232
023A
0238
0218
021A
0212
0210
0410
0412
041A
0418
0438
043A
0432
0430
04B0
04B2
04BA
04B8
0498
049A
0492
0490
0690
0692
069A
0698
06B8
06BA
06B2
06B0
0630
0632
063A
0638
0618
061A
0612
0610
13 15
02D0
02D2
02DA
02D8
02F8
02FA
02F2
02F0
0270
0272
027A
0278
0258
025A
0252
0250
0450
0452
045A
0458
0478
047A
0472
0470
04F0
04F2
04FA
04F8
04D8
04DA
04D2
04D0
06D0
06D2
06DA
06D8
06F8
06FA
06F2
06F0
0670
0672
067A
0678
0658
065A
0652
0650
17
02D1 02D5 02D4 02C4 02C5 02C1

02D3 02D7 02D6 02C6 02C7 02C3
02DB 02DF 02DE 02CE 02CF 02CB
02D9 02DD 02DC 02CC 02CD 02C9
02F9 02FD 02FC 02EC 02ED 02E9
02FB 02FF 02FE 02EE 02EF 02EB
02F3 02F7 02F6 02E6 02E7 02E3
02F1 02F5 02F4 02E4 02E5 02E1
0271 0275 0274 0264 0265 0261
0273 0277 0276 0266 0267 0263
027B 027F 027E 026E 026F 026B
0279 027D 027C 026C 026D 0269
0259 025D 025C 024C 024D 0249
025B 025F 025E 024E 024F 024B
0253 0257 0256 0246 0247 0243
0251 0255 0254 0244 0245 0241
0451 0455 0454 0444 0445 0441
0453 0457 0456 0446 0447 0443
045B 045F 045E 044E 044F 044B
0459 045D 045C 044C 044D 0449
0479 047D 047C 046C 046D 0469
047B 047F 047E 046E 046F 046B
0473 0477 0476 0466 0467 0463
0471 0475 0474 0464 0465 0461
04F1 04F5 04F4 04E4 04E5 04E1
04F3 04F7 04F6 04E6 04E7 04E3
04F9 04FD 04FC 04EC 04ED 04E9
04D9 04DD 04DC 04CC 04CD 04C9
04D3 04D7 04D6 04C6 04C7 04C3
04D1 04D5 04D4 04C4 04C5 04C1
06D1 06D5 06D4 06C4 06C5 06C1
06D3 06D7 06D6 06C6 06C7 06C3
06D9 06DD 06DC 06CC 06CD 06C9
06F9 06FD 06FC 06EC 06ED 06E9
06F3 06F7 06F6 06E6 06E7 06E3
06F1 06F5 06F4 06E4 06E5 06E1
0671 0675 0674 0664 0665 0661
0673 0677 0676 0666 0667 0663
067B 067F 067E 066E 066F 066B
0679 067D 067C 066C 066D 0669
0659 065D 065C 064C 064D 0649
065B 065F 065E 064E 064F 064B
0653 0657 0656 0646 0647 0643
0651 0655 0654 0644 0645 0641
19 21
23
25
27 29
02C0
02C2
02CA
02C8
02E8
02EA
02E2
02E0
0260
0262
026A
0268
0248
024A
0242
0240
0440
0442
044A
0448
0468
046A
0462
0460
04E0
04E2
04EA
04E8
04C8
04CA
04C2
04C0
06C0
06C2
06CA
06C8
06E8
06EA
06E2
06E0
0660
0662
066A
0668
0648
064A
0642
0640
31
00C0 00C1
00C2 00C3
00CA 00CB
00C8 00C9
00E8 00E9
00EA 00EB
00E2 00E3
00E0 00E1
0060 0061
0062 0063
006A 006B
0068 0069
0048 0049
004A 004B
0042 0043
0040 0041
0540 0541
0542 0543
054A 054B
0548 0549
0568 0569
056A 056B
0562 0563
0560 0561
05E0 05E1
05E2 05E3
05EA 05EB
05E8 05E9
05C8 05C9
05CA 05CB
05C2 05C3
05C0 05C1
07C0 07C1
07C2 07C3
07CA 07CB
07C8 07C9
07E8 07E9
07EA 07EB
07E2 07E3
07E0 07E1
0760 0761
0762 0763
076A 076B
0768 0769
0748 0749
074A 074B
0742 0743
0740 0741
33 35
00C5 00C4 00D4 00D5 00D1

00C7 00C6 00D6 00D7 00D3
00CF 00CE 00DE 00DF 00DB
00CD 00CC 00DC 00DD 00D9
00ED 00EC 00FC 00FD 00F9
00EF 00EE 00FE 00FF 00FB
00E7 00E6 00F6 00F7 00F3
00E5 00E4 00F4 00F5 00F1
0065 0064 0074 0075 0071
0067 0066 0076 0077 0073
006F 006E 007E 007F 007B
006D 006C 007C 007D 0079
004D 004C 005C 005D 0059
004F 004E 005E 005F 005B
0047 0046 0056 0057 0053
0045 0044 0054 0055 0051
0545 0544 0554 0555 0551
0547 0546 0556 0557 0553
054F 054E 055E 055F 055B
054D 054C 055C 055D 0559
056D 056C 057C 057D 0579
056F 056E 057E 057F 057B
0567 0566 0576 0577 0573
0565 0564 0574 0575 0571
05E5 05E4 05F4 05F5 05F1
05E7 05E6 05F6 05F7 05F3
05ED 05EC 05FC 05FD 05F9
05CD 05CC 05DC 05DD 05D9
05C7 05C6 05D6 05D7 05D3
05C5 05C4 05D4 05D5 05D1
07C5 07C4 07D4 07D5 07D1
07C7 07C6 07D6 07D7 07D3
07CD 07CC 07DC 07DD 07D9
07ED 07EC 07FC 07FD 07F9
07E7 07E6 07F6 07F7 07F3
07E5 07E4 07F4 07F5 07F1
0765 0764 0774 0775 0771
0767 0766 0776 0777 0773
076F 076E 077E 077F 077B
076D 076C 077C 077D 0779
074D 074C 075C 075D 0759
074F 074E 075E 075F 075B
0747 0746 0756 0757 0753
0745 0744 0754 0755 0751
37
39
41 43
45
00D0 0090
00D2 0092
00DA 009A
00D8 0098
00F8 00B8
00FA 00BA
00F2 00B2
00F0 00B0
0070 0030
0072 0032
007A 003A
0078 0038
0058 0018
005A 001A
0052 0012
0050 0010
0550 0510
0552 0512
055A 051A
0558 0518
0578 0538
057A 053A
0572 0532
0570 0530
05F0 05B0
05F2 05B2
05FA 05BA
05F8 05B8
05D8 0598
05DA 059A
05D2 0592
05D0 0590
07D0 0790
07D2 0792
07DA 079A
07D8 0798
07F8 07B8
07FA 07BA
07F2 07B2
07F0 07B0
0770 0730
0772 0732
077A 073A
0778 0738
0758 0718
075A 071A
0752 0712
0750 0710
47 49
0091
0093
009B
0099
00B9
00BB
00B3
00B1
0031
0033
003B
0039
0019
001B
0013
0011
0511
0513
051B
0519
0539
053B
0533
0531
05B1
05B3
05BB
05B9
0599
059B
0593
0591
0791
0793
079B
0799
07B9
07BB
07B3
07B1
0731
0733
073B
0739
0719
071B
0713
0711
51
0095 0094
0097 0096
009F 009E
009D 009C
00BD 00BC
00BF 00BE
00B7 00B6
00B5 00B4
0035 0034
0037 0036
003F 003E
003D 003C
001D 001C
001F 001E
0017 0016
0015 0014
0515 0514
0517 0516
051F 051E
051D 051C
053D 053C
053F 053E
0537 0536
0535 0534
05B5 05B4
05B7 05B6
05BF 05BE
05BD 05BC
059D 059C
059F 059E
0597 0596
0595 0594
0795 0794
0797 0796
079F 079E
079D 079C
07BD 07BC
07BF 07BE
07B7 07B6
07B5 07B4
0735 0734
0737 0736
073F 073E
073D 073C
071D 071C
071F 071E
0717 0716
0715 0714
53 55
0084
0086
008E
008C
00AC
00AE
00A6
00A4
0024
0026
002E
002C
000C
000E
0006
0004
0504
0506
050E
050C
052C
052E
0526
0524
05A4
05A6
05AE
05AC
058C
058E
0586
0584
0784
0786
078E
078C
07AC
07AE
07A6
07A4
0724
0726
072E
072C
070C
070E
0706
0704
57
0085
0087
008F
008D
00AD
00AF
00A7
00A5
0025
0027
002F
002D
000D
000F
0007
0005
0505
0507
050F
050D
052D
052F
0527
0525
05A5
05A7
05AF
05AD
058D
058F
0587
0585
0785
0787
078F
078D
07AD
07AF
07A7
07A5
0725
0727
072F
072D
070D
070F
0707
0705
59
0081
0083
008B
0089
00A9
00AB
00A3
00A1
0021
0023
002B
0029
0009
000B
0003
0001
0501
0503
050B
0509
0529
052B
0523
0521
05A1
05A3
05AB
05A9
0589
058B
0583
0581
0781
0783
078B
0789
07A9
07AB
07A3
07A1
0721
0723
072B
0729
0709
070B
0703
0701
0080
0082
008A
0088
00A8
00AA
00A2
00A0
0020
0022
002A
0028
0008
000A
0002
0000
0500
0502
050A
0508
0528
052A
0522
0520
05A0
05A2
05AA
05A8
0588
058A
0582
0580
0780
0782
078A
0788
07A8
07AA
07A2
07A0
0720
0722
072A
0728
0708
070A
0702
0700
61 63
0100
0102
010A
0108
0128
012A
0122
0120
01A0
01A2
01AA
01A8
0188
018A
0182
0180
0380
0382
038A
0388
03A8
03AA
03A2
03A0
0320
0322
032A
0328
0308
030A
0302
0300
65
0101
0103
010B
0109
0129
012B
0123
0121
01A1
01A3
01AB
01A9
0189
018B
0183
0181
0381
0383
038B
0389
03A9
03AB
03A3
03A1
0321
0323
032B
0329
0309
030B
0303
0301
0105
0107
010F
010D
012D
012F
0127
0125
01A5
01A7
01AF
01AD
018D
018F
0187
0185
0385
0387
038F
038D
03AD
03AF
03A7
03A5
0325
0327
032F
032D
030D
030F
0307
0305
67 69
0104
0106
010E
010C
012C
012E
0126
0124
01A4
01A6
01AE
01AC
018C
018E
0186
0184
0384
0386
038E
038C
03AC
03AE
03A6
03A4
0324
0326
032E
032C
030C
030E
0306
0304
71
0114
0116
011E
011C
013C
013E
0136
0134
01B4
01B6
01BE
01BC
019C
019E
0196
0194
0394
0396
039E
039C
03BC
03BE
03B6
03B4
0334
0336
033E
033C
031C
031E
0316
0314
73
0115 0111
0117 0113
011F 011B
011D 0119
013D 0139
013F 013B
0137 0133
0135 0131
01B5 01B1
01B7 01B3
01BF 01BB
01BD 01B9
019D 0199
019F 019B
0197 0193
0195 0191
0395 0391
0397 0393
039F 039B
039D 0399
03BD 03B9
03BF 03BB
03B7 03B3
03B5 03B1
0335 0331
0337 0333
033F 033B
033D 0339
031D 0319
031F 031B
0317 0313
0315 0311
75 77
0110
0112
011A
0118
0138
013A
0132
0130
01B0
01B2
01BA
01B8
0198
019A
0192
0190
0390
0392
039A
0398
03B8
03BA
03B2
03B0
0330
0332
033A
0338
0318
031A
0312
0310
79
0150 0151
0152 0153
015A 015B
0158 0159
0178 0179
017A 017B
0172 0173
0170 0171
01F0 01F1
01F2 01F3
01FA 01FB
01F8 01F9
01D8 01D9
01DA 01DB
01D2 01D3
01D0 01D1
03D0 03D1
03D2 03D3
03DA 03DB
03D8 03D9
03F8 03F9
03FA 03FB
03F2 03F3
03F0 03F1
0370 0371
0372 0373
037A 037B
0378 0379
0358 0359
035A 035B
0352 0353
0350 0351
81 83
0155 0154 0144 0145 0141

0157 0156 0146 0147 0143
015F 015E 014E 014F 014B
015D 015C 014C 014D 0149
017D 017C 016C 016D 0169
017F 017E 016E 016F 016B
0177 0176 0166 0167 0163
0175 0174 0164 0165 0161
01F5 01F4 01E4 01E5 01E1
01F7 01F6 01E6 01E7 01E3
01FF 01FE 01EE 01EF 01EB
01FD 01FC 01EC 01ED 01E9
01DD 01DC 01CC 01CD 01C9
01DF 01DE 01CE 01CF 01CB
01D7 01D6 01C6 01C7 01C3
01D5 01D4 01C4 01C5 01C1
03D5 03D4 03C4 03C5 03C1
03D7 03D6 03C6 03C7 03C3
03DF 03DE 03CE 03CF 03CB
03DD 03DC 03CC 03CD 03C9
03FD 03FC 03EC 03ED 03E9
03FF 03FE 03EE 03EF 03EB
03F7 03F6 03E6 03E7 03E3
03F5 03F4 03E4 03E5 03E1
0375 0374 0364 0365 0361
0377 0376 0366 0367 0363
037F 037E 036E 036F 036B
037D 037C 036C 036D 0369
035D 035C 034C 034D 0349
035F 035E 034E 034F 034B
0357 0356 0346 0347 0343
0355 0354 0344 0345 0341
85
87
89 91
93
0140
0142
014A
0148
0168
016A
0162
0160
01E0
01E2
01EA
01E8
01C8
01CA
01C2
01C0
03C0
03C2
03CA
03C8
03E8
03EA
03E2
03E0
0360
0362
036A
0368
0348
034A
0342
0340
95
194
CableLabs
10/29/13
127
125
123
121
119
117
115
113
111
109
107
105
103
101
99
97
95
93
91
89
87
85
83
81
79
77
75
73
71
69
67
65
63
61
59
57
55
53
51
49
47
45
43
41
39
37
35
33
31
29
27
25
23
21
19
17
15
13
11
9
7
5
3
1
CM-SP-PHYv3.1-I01-131029
0800 0802 080A 0808 0828 082A 0822 0820 08A0 08A2 08AA 08A8 0888 088A 0882 0880 0A80 0A82 0A8A 0A88 0AA8 0AAA 0AA2 0AA0 0A20 0A22 0A2A 0A28 0A08 0A0A 0A02 0A00 0200 0202 020A 0208 0228 022A 0222 0220 02A0 02A2 02AA 02A8 0288 028A 0282 0280 0080 0082 008A 0088 00A8 00AA 00A2 00A0 0020 0022 002A 0028 0008 000A 0002 0000
0801 0803 080B 0809 0829 082B 0823 0821 08A1 08A3 08AB 08A9 0889 088B 0883 0881 0A81 0A83 0A8B 0A89 0AA9 0AAB 0AA3 0AA1 0A21 0A23 0A2B 0A29 0A09 0A0B 0A03 0A01 0201 0203 020B 0209 0229 022B 0223 0221 02A1 02A3 02AB 02A9 0289 028B 0283 0281 0081 0083 008B 0089 00A9 00AB 00A3 00A1 0021 0023 002B 0029 0009 000B 0003 0001
0805 0807 080F 080D 082D 082F 0827 0825 08A5 08A7 08AF 08AD 088D 088F 0887 0885 0A85 0A87 0A8F 0A8D 0AAD 0AAF 0AA7 0AA5 0A25 0A27 0A2F 0A2D 0A0D 0A0F 0A07 0A05 0205 0207 020F 020D 022D 022F 0227 0225 02A5 02A7 02AF 02AD 028D 028F 0287 0285 0085 0087 008F 008D 00AD 00AF 00A7 00A5 0025 0027 002F 002D 000D 000F 0007 0005
0804 0806 080E 080C 082C 082E 0826 0824 08A4 08A6 08AE 08AC 088C 088E 0886 0884 0A84 0A86 0A8E 0A8C 0AAC 0AAE 0AA6 0AA4 0A24 0A26 0A2E 0A2C 0A0C 0A0E 0A06 0A04 0204 0206 020E 020C 022C 022E 0226 0224 02A4 02A6 02AE 02AC 028C 028E 0286 0284 0084 0086 008E 008C 00AC 00AE 00A6 00A4 0024 0026 002E 002C 000C 000E 0006 0004
0814 0816 081E 081C 083C 083E 0836 0834 08B4 08B6 08BE 08BC 089C 089E 0896 0894 0A94 0A96 0A9E 0A9C 0ABC 0ABE 0AB6 0AB4 0A34 0A36 0A3E 0A3C 0A1C 0A1E 0A16 0A14 0214 0216 021E 021C 023C 023E 0236 0234 02B4 02B6 02BE 02BC 029C 029E 0296 0294 0094 0096 009E 009C 00BC 00BE 00B6 00B4 0034 0036 003E 003C 001C 001E 0016 0014
0815 0817 081F 081D 083D 083F 0837 0835 08B5 08B7 08BF 08BD 089D 089F 0897 0895 0A95 0A97 0A9F 0A9D 0ABD 0ABF 0AB7 0AB5 0A35 0A37 0A3F 0A3D 0A1D 0A1F 0A17 0A15 0215 0217 021F 021D 023D 023F 0237 0235 02B5 02B7 02BF 02BD 029D 029F 0297 0295 0095 0097 009F 009D 00BD 00BF 00B7 00B5 0035 0037 003F 003D 001D 001F 0017 0015
0811 0813 081B 0819 0839 083B 0833 0831 08B1 08B3 08BB 08B9 0899 089B 0893 0891 0A91 0A93 0A9B 0A99 0AB9 0ABB 0AB3 0AB1 0A31 0A33 0A3B 0A39 0A19 0A1B 0A13 0A11 0211 0213 021B 0219 0239 023B 0233 0231 02B1 02B3 02BB 02B9 0299 029B 0293 0291 0091 0093 009B 0099 00B9 00BB 00B3 00B1 0031 0033 003B 0039 0019 001B 0013 0011
0810 0812 081A 0818 0838 083A 0832 0830 08B0 08B2 08BA 08B8 0898 089A 0892 0890 0A90 0A92 0A9A 0A98 0AB8 0ABA 0AB2 0AB0 0A30 0A32 0A3A 0A38 0A18 0A1A 0A12 0A10 0210 0212 021A 0218 0238 023A 0232 0230 02B0 02B2 02BA 02B8 0298 029A 0292 0290 0090 0092 009A 0098 00B8 00BA 00B2 00B0 0030 0032 003A 0038 0018 001A 0012 0010
0850 0852 085A 0858 0878 087A 0872 0870 08F0 08F2 08FA 08F8 08D8 08DA 08D2 08D0 0AD0 0AD2 0ADA 0AD8 0AF8 0AFA 0AF2 0AF0 0A70 0A72 0A7A 0A78 0A58 0A5A 0A52 0A50 0250 0252 025A 0258 0278 027A 0272 0270 02F0 02F2 02FA 02F8 02D8 02DA 02D2 02D0 00D0 00D2 00DA 00D8 00F8 00FA 00F2 00F0 0070 0072 007A 0078 0058 005A 0052 0050
0851 0853 085B 0859 0879 087B 0873 0871 08F1 08F3 08FB 08F9 08D9 08DB 08D3 08D1 0AD1 0AD3 0ADB 0AD9 0AF9 0AFB 0AF3 0AF1 0A71 0A73 0A7B 0A79 0A59 0A5B 0A53 0A51 0251 0253 025B 0259 0279 027B 0273 0271 02F1 02F3 02FB 02F9 02D9 02DB 02D3 02D1 00D1 00D3 00DB 00D9 00F9 00FB 00F3 00F1 0071 0073 007B 0079 0059 005B 0053 0051
0855 0857 085F 085D 087D 087F 0877 0875 08F5 08F7 08FF 08FD 08DD 08DF 08D7 08D5 0AD5 0AD7 0ADF 0ADD 0AFD 0AFF 0AF7 0AF5 0A75 0A77 0A7F 0A7D 0A5D 0A5F 0A57 0A55 0255 0257 025F 025D 027D 027F 0277 0275 02F5 02F7 02FF 02FD 02DD 02DF 02D7 02D5 00D5 00D7 00DF 00DD 00FD 00FF 00F7 00F5 0075 0077 007F 007D 005D 005F 0057 0055
0854 0856 085E 085C 087C 087E 0876 0874 08F4 08F6 08FE 08FC 08DC 08DE 08D6 08D4 0AD4 0AD6 0ADE 0ADC 0AFC 0AFE 0AF6 0AF4 0A74 0A76 0A7E 0A7C 0A5C 0A5E 0A56 0A54 0254 0256 025E 025C 027C 027E 0276 0274 02F4 02F6 02FE 02FC 02DC 02DE 02D6 02D4 00D4 00D6 00DE 00DC 00FC 00FE 00F6 00F4 0074 0076 007E 007C 005C 005E 0056 0054
0844 0846 084E 084C 086C 086E 0866 0864 08E4 08E6 08EE 08EC 08CC 08CE 08C6 08C4 0AC4 0AC6 0ACE 0ACC 0AEC 0AEE 0AE6 0AE4 0A64 0A66 0A6E 0A6C 0A4C 0A4E 0A46 0A44 0244 0246 024E 024C 026C 026E 0266 0264 02E4 02E6 02EE 02EC 02CC 02CE 02C6 02C4 00C4 00C6 00CE 00CC 00EC 00EE 00E6 00E4 0064 0066 006E 006C 004C 004E 0046 0044
0845 0847 084F 084D 086D 086F 0867 0865 08E5 08E7 08EF 08ED 08CD 08CF 08C7 08C5 0AC5 0AC7 0ACF 0ACD 0AED 0AEF 0AE7 0AE5 0A65 0A67 0A6F 0A6D 0A4D 0A4F 0A47 0A45 0245 0247 024F 024D 026D 026F 0267 0265 02E5 02E7 02EF 02ED 02CD 02CF 02C7 02C5 00C5 00C7 00CF 00CD 00ED 00EF 00E7 00E5 0065 0067 006F 006D 004D 004F 0047 0045
0841 0843 084B 0849 0869 086B 0863 0861 08E1 08E3 08EB 08E9 08C9 08CB 08C3 08C1 0AC1 0AC3 0ACB 0AC9 0AE9 0AEB 0AE3 0AE1 0A61 0A63 0A6B 0A69 0A49 0A4B 0A43 0A41 0241 0243 024B 0249 0269 026B 0263 0261 02E1 02E3 02EB 02E9 02C9 02CB 02C3 02C1 00C1 00C3 00CB 00C9 00E9 00EB 00E3 00E1 0061 0063 006B 0069 0049 004B 0043 0041
0840 0842 084A 0848 0868 086A 0862 0860 08E0 08E2 08EA 08E8 08C8 08CA 08C2 08C0 0AC0 0AC2 0ACA 0AC8 0AE8 0AEA 0AE2 0AE0 0A60 0A62 0A6A 0A68 0A48 0A4A 0A42 0A40 0240 0242 024A 0248 0268 026A 0262 0260 02E0 02E2 02EA 02E8 02C8 02CA 02C2 02C0 00C0 00C2 00CA 00C8 00E8 00EA 00E2 00E0 0060 0062 006A 0068 0048 004A 0042 0040
0940 0942 094A 0948 0968 096A 0962 0960 09E0 09E2 09EA 09E8 09C8 09CA 09C2 09C0 0BC0 0BC2 0BCA 0BC8 0BE8 0BEA 0BE2 0BE0 0B60 0B62 0B6A 0B68 0B48 0B4A 0B42 0B40 0340 0342 034A 0348 0368 036A 0362 0360 03E0 03E2 03EA 03E8 03C8 03CA 03C2 03C0 01C0 01C2 01CA 01C8 01E8 01EA 01E2 01E0 0160 0162 016A 0168 0148 014A 0142 0140
0941 0943 094B 0949 0969 096B 0963 0961 09E1 09E3 09EB 09E9 09C9 09CB 09C3 09C1 0BC1 0BC3 0BCB 0BC9 0BE9 0BEB 0BE3 0BE1 0B61 0B63 0B6B 0B69 0B49 0B4B 0B43 0B41 0341 0343 034B 0349 0369 036B 0363 0361 03E1 03E3 03EB 03E9 03C9 03CB 03C3 03C1 01C1 01C3 01CB 01C9 01E9 01EB 01E3 01E1 0161 0163 016B 0169 0149 014B 0143 0141
0945 0947 094F 094D 096D 096F 0967 0965 09E5 09E7 09EF 09ED 09CD 09CF 09C7 09C5 0BC5 0BC7 0BCF 0BCD 0BED 0BEF 0BE7 0BE5 0B65 0B67 0B6F 0B6D 0B4D 0B4F 0B47 0B45 0345 0347 034F 034D 036D 036F 0367 0365 03E5 03E7 03EF 03ED 03CD 03CF 03C7 03C5 01C5 01C7 01CF 01CD 01ED 01EF 01E7 01E5 0165 0167 016F 016D 014D 014F 0147 0145
0944 0946 094E 094C 096C 096E 0966 0964 09E4 09E6 09EE 09EC 09CC 09CE 09C6 09C4 0BC4 0BC6 0BCE 0BCC 0BEC 0BEE 0BE6 0BE4 0B64 0B66 0B6E 0B6C 0B4C 0B4E 0B46 0B44 0344 0346 034E 034C 036C 036E 0366 0364 03E4 03E6 03EE 03EC 03CC 03CE 03C6 03C4 01C4 01C6 01CE 01CC 01EC 01EE 01E6 01E4 0164 0166 016E 016C 014C 014E 0146 0144
0954 0956 095E 095C 097C 097E 0976 0974 09F4 09F6 09FE 09FC 09DC 09DE 09D6 09D4 0BD4 0BD6 0BDE 0BDC 0BFC 0BFE 0BF6 0BF4 0B74 0B76 0B7E 0B7C 0B5C 0B5E 0B56 0B54 0354 0356 035E 035C 037C 037E 0376 0374 03F4 03F6 03FE 03FC 03DC 03DE 03D6 03D4 01D4 01D6 01DE 01DC 01FC 01FE 01F6 01F4 0174 0176 017E 017C 015C 015E 0156 0154
0955 0957 095F 095D 097D 097F 0977 0975 09F5 09F7 09FF 09FD 09DD 09DF 09D7 09D5 0BD5 0BD7 0BDF 0BDD 0BFD 0BFF 0BF7 0BF5 0B75 0B77 0B7F 0B7D 0B5D 0B5F 0B57 0B55 0355 0357 035F 035D 037D 037F 0377 0375 03F5 03F7 03FF 03FD 03DD 03DF 03D7 03D5 01D5 01D7 01DF 01DD 01FD 01FF 01F7 01F5 0175 0177 017F 017D 015D 015F 0157 0155
0951 0953 095B 0959 0979 097B 0973 0971 09F1 09F3 09FB 09F9 09D9 09DB 09D3 09D1 0BD1 0BD3 0BDB 0BD9 0BF9 0BFB 0BF3 0BF1 0B71 0B73 0B7B 0B79 0B59 0B5B 0B53 0B51 0351 0353 035B 0359 0379 037B 0373 0371 03F1 03F3 03FB 03F9 03D9 03DB 03D3 03D1 01D1 01D3 01DB 01D9 01F9 01FB 01F3 01F1 0171 0173 017B 0179 0159 015B 0153 0151
0950 0952 095A 0958 0978 097A 0972 0970 09F0 09F2 09FA 09F8 09D8 09DA 09D2 09D0 0BD0 0BD2 0BDA 0BD8 0BF8 0BFA 0BF2 0BF0 0B70 0B72 0B7A 0B78 0B58 0B5A 0B52 0B50 0350 0352 035A 0358 0378 037A 0372 0370 03F0 03F2 03FA 03F8 03D8 03DA 03D2 03D0 01D0 01D2 01DA 01D8 01F8 01FA 01F2 01F0 0170 0172 017A 0178 0158 015A 0152 0150
0910 0912 091A 0918 0938 093A 0932 0930 09B0 09B2 09BA 09B8 0998 099A 0992 0990 0B90 0B92 0B9A 0B98 0BB8 0BBA 0BB2 0BB0 0B30 0B32 0B3A 0B38 0B18 0B1A 0B12 0B10 0310 0312 031A 0318 0338 033A 0332 0330 03B0 03B2 03BA 03B8 0398 039A 0392 0390 0190 0192 019A 0198 01B8 01BA 01B2 01B0 0130 0132 013A 0138 0118 011A 0112 0110
0911 0913 091B 0919 0939 093B 0933 0931 09B1 09B3 09BB 09B9 0999 099B 0993 0991 0B91 0B93 0B9B 0B99 0BB9 0BBB 0BB3 0BB1 0B31 0B33 0B3B 0B39 0B19 0B1B 0B13 0B11 0311 0313 031B 0319 0339 033B 0333 0331 03B1 03B3 03BB 03B9 0399 039B 0393 0391 0191 0193 019B 0199 01B9 01BB 01B3 01B1 0131 0133 013B 0139 0119 011B 0113 0111
0915 0917 091F 091D 093D 093F 0937 0935 09B5 09B7 09BF 09BD 099D 099F 0997 0995 0B95 0B97 0B9F 0B9D 0BBD 0BBF 0BB7 0BB5 0B35 0B37 0B3F 0B3D 0B1D 0B1F 0B17 0B15 0315 0317 031F 031D 033D 033F 0337 0335 03B5 03B7 03BF 03BD 039D 039F 0397 0395 0195 0197 019F 019D 01BD 01BF 01B7 01B5 0135 0137 013F 013D 011D 011F 0117 0115
0914 0916 091E 091C 093C 093E 0936 0934 09B4 09B6 09BE 09BC 099C 099E 0996 0994 0B94 0B96 0B9E 0B9C 0BBC 0BBE 0BB6 0BB4 0B34 0B36 0B3E 0B3C 0B1C 0B1E 0B16 0B14 0314 0316 031E 031C 033C 033E 0336 0334 03B4 03B6 03BE 03BC 039C 039E 0396 0394 0194 0196 019E 019C 01BC 01BE 01B6 01B4 0134 0136 013E 013C 011C 011E 0116 0114
0904 0906 090E 090C 092C 092E 0926 0924 09A4 09A6 09AE 09AC 098C 098E 0986 0984 0B84 0B86 0B8E 0B8C 0BAC 0BAE 0BA6 0BA4 0B24 0B26 0B2E 0B2C 0B0C 0B0E 0B06 0B04 0304 0306 030E 030C 032C 032E 0326 0324 03A4 03A6 03AE 03AC 038C 038E 0386 0384 0184 0186 018E 018C 01AC 01AE 01A6 01A4 0124 0126 012E 012C 010C 010E 0106 0104
0905 0907 090F 090D 092D 092F 0927 0925 09A5 09A7 09AF 09AD 098D 098F 0987 0985 0B85 0B87 0B8F 0B8D 0BAD 0BAF 0BA7 0BA5 0B25 0B27 0B2F 0B2D 0B0D 0B0F 0B07 0B05 0305 0307 030F 030D 032D 032F 0327 0325 03A5 03A7 03AF 03AD 038D 038F 0387 0385 0185 0187 018F 018D 01AD 01AF 01A7 01A5 0125 0127 012F 012D 010D 010F 0107 0105
0901 0903 090B 0909 0929 092B 0923 0921 09A1 09A3 09AB 09A9 0989 098B 0983 0981 0B81 0B83 0B8B 0B89 0BA9 0BAB 0BA3 0BA1 0B21 0B23 0B2B 0B29 0B09 0B0B 0B03 0B01 0301 0303 030B 0309 0329 032B 0323 0321 03A1 03A3 03AB 03A9 0389 038B 0383 0381 0181 0183 018B 0189 01A9 01AB 01A3 01A1 0121 0123 012B 0129 0109 010B 0103 0101
0900 0902 090A 0908 0928 092A 0922 0920 09A0 09A2 09AA 09A8 0988 098A 0982 0980 0B80 0B82 0B8A 0B88 0BA8 0BAA 0BA2 0BA0 0B20 0B22 0B2A 0B28 0B08 0B0A 0B02 0B00 0300 0302 030A 0308 0328 032A 0322 0320 03A0 03A2 03AA 03A8 0388 038A 0382 0380 0180 0182 018A 0188 01A8 01AA 01A2 01A0 0120 0122 012A 0128 0108 010A 0102 0100
0D00 0D02 0D0A 0D08 0D28 0D2A 0D22 0D20 0DA0 0DA2 0DAA 0DA8 0D88 0D8A 0D82 0D80 0F80 0F82 0F8A 0F88 0FA8 0FAA 0FA2 0FA0 0F20 0F22 0F2A 0F28 0F08 0F0A 0F02 0F00 0700 0702 070A 0708 0728 072A 0722 0720 07A0 07A2 07AA 07A8 0788 078A 0782 0780 0580 0582 058A 0588 05A8 05AA 05A2 05A0 0520 0522 052A 0528 0508 050A 0502 0500
0D01 0D03 0D0B 0D09 0D29 0D2B 0D23 0D21 0DA1 0DA3 0DAB 0DA9 0D89 0D8B 0D83 0D81 0F81 0F83 0F8B 0F89 0FA9 0FAB 0FA3 0FA1 0F21 0F23 0F2B 0F29 0F09 0F0B 0F03 0F01 0701 0703 070B 0709 0729 072B 0723 0721 07A1 07A3 07AB 07A9 0789 078B 0783 0781 0581 0583 058B 0589 05A9 05AB 05A3 05A1 0521 0523 052B 0529 0509 050B 0503 0501
0D05 0D07 0D0F 0D0D 0D2D 0D2F 0D27 0D25 0DA5 0DA7 0DAF 0DAD 0D8D 0D8F 0D87 0D85 0F85 0F87 0F8F 0F8D 0FAD 0FAF 0FA7 0FA5 0F25 0F27 0F2F 0F2D 0F0D 0F0F 0F07 0F05 0705 0707 070F 070D 072D 072F 0727 0725 07A5 07A7 07AF 07AD 078D 078F 0787 0785 0585 0587 058F 058D 05AD 05AF 05A7 05A5 0525 0527 052F 052D 050D 050F 0507 0505
0D04 0D06 0D0E 0D0C 0D2C 0D2E 0D26 0D24 0DA4 0DA6 0DAE 0DAC 0D8C 0D8E 0D86 0D84 0F84 0F86 0F8E 0F8C 0FAC 0FAE 0FA6 0FA4 0F24 0F26 0F2E 0F2C 0F0C 0F0E 0F06 0F04 0704 0706 070E 070C 072C 072E 0726 0724 07A4 07A6 07AE 07AC 078C 078E 0786 0784 0584 0586 058E 058C 05AC 05AE 05A6 05A4 0524 0526 052E 052C 050C 050E 0506 0504
0D14 0D16 0D1E 0D1C 0D3C 0D3E 0D36 0D34 0DB4 0DB6 0DBE 0DBC 0D9C 0D9E 0D96 0D94 0F94 0F96 0F9E 0F9C 0FBC 0FBE 0FB6 0FB4 0F34 0F36 0F3E 0F3C 0F1C 0F1E 0F16 0F14 0714 0716 071E 071C 073C 073E 0736 0734 07B4 07B6 07BE 07BC 079C 079E 0796 0794 0594 0596 059E 059C 05BC 05BE 05B6 05B4 0534 0536 053E 053C 051C 051E 0516 0514
0D15 0D17 0D1F 0D1D 0D3D 0D3F 0D37 0D35 0DB5 0DB7 0DBF 0DBD 0D9D 0D9F 0D97 0D95 0F95 0F97 0F9F 0F9D 0FBD 0FBF 0FB7 0FB5 0F35 0F37 0F3F 0F3D 0F1D 0F1F 0F17 0F15 0715 0717 071F 071D 073D 073F 0737 0735 07B5 07B7 07BF 07BD 079D 079F 0797 0795 0595 0597 059F 059D 05BD 05BF 05B7 05B5 0535 0537 053F 053D 051D 051F 0517 0515
0D11 0D13 0D1B 0D19 0D39 0D3B 0D33 0D31 0DB1 0DB3 0DBB 0DB9 0D99 0D9B 0D93 0D91 0F91 0F93 0F9B 0F99 0FB9 0FBB 0FB3 0FB1 0F31 0F33 0F3B 0F39 0F19 0F1B 0F13 0F11 0711 0713 071B 0719 0739 073B 0733 0731 07B1 07B3 07BB 07B9 0799 079B 0793 0791 0591 0593 059B 0599 05B9 05BB 05B3 05B1 0531 0533 053B 0539 0519 051B 0513 0511
0D10 0D12 0D1A 0D18 0D38 0D3A 0D32 0D30 0DB0 0DB2 0DBA 0DB8 0D98 0D9A 0D92 0D90 0F90 0F92 0F9A 0F98 0FB8 0FBA 0FB2 0FB0 0F30 0F32 0F3A 0F38 0F18 0F1A 0F12 0F10 0710 0712 071A 0718 0738 073A 0732 0730 07B0 07B2 07BA 07B8 0798 079A 0792 0790 0590 0592 059A 0598 05B8 05BA 05B2 05B0 0530 0532 053A 0538 0518 051A 0512 0510
0D50 0D52 0D5A 0D58 0D78 0D7A 0D72 0D70 0DF0 0DF2 0DFA 0DF8 0DD8 0DDA 0DD2 0DD0 0FD0 0FD2 0FDA 0FD8 0FF8 0FFA 0FF2 0FF0 0F70 0F72 0F7A 0F78 0F58 0F5A 0F52 0F50 0750 0752 075A 0758 0778 077A 0772 0770 07F0 07F2 07FA 07F8 07D8 07DA 07D2 07D0 05D0 05D2 05DA 05D8 05F8 05FA 05F2 05F0 0570 0572 057A 0578 0558 055A 0552 0550
0D51 0D53 0D5B 0D59 0D79 0D7B 0D73 0D71 0DF1 0DF3 0DFB 0DF9 0DD9 0DDB 0DD3 0DD1 0FD1 0FD3 0FDB 0FD9 0FF9 0FFB 0FF3 0FF1 0F71 0F73 0F7B 0F79 0F59 0F5B 0F53 0F51 0751 0753 075B 0759 0779 077B 0773 0771 07F1 07F3 07FB 07F9 07D9 07DB 07D3 07D1 05D1 05D3 05DB 05D9 05F9 05FB 05F3 05F1 0571 0573 057B 0579 0559 055B 0553 0551
0D55 0D57 0D5F 0D5D 0D7D 0D7F 0D77 0D75 0DF5 0DF7 0DFF 0DFD 0DDD 0DDF 0DD7 0DD5 0FD5 0FD7 0FDF 0FDD 0FFD 0FFF 0FF7 0FF5 0F75 0F77 0F7F 0F7D 0F5D 0F5F 0F57 0F55 0755 0757 075F 075D 077D 077F 0777 0775 07F5 07F7 07FF 07FD 07DD 07DF 07D7 07D5 05D5 05D7 05DF 05DD 05FD 05FF 05F7 05F5 0575 0577 057F 057D 055D 055F 0557 0555
0D54 0D56 0D5E 0D5C 0D7C 0D7E 0D76 0D74 0DF4 0DF6 0DFE 0DFC 0DDC 0DDE 0DD6 0DD4 0FD4 0FD6 0FDE 0FDC 0FFC 0FFE 0FF6 0FF4 0F74 0F76 0F7E 0F7C 0F5C 0F5E 0F56 0F54 0754 0756 075E 075C 077C 077E 0776 0774 07F4 07F6 07FE 07FC 07DC 07DE 07D6 07D4 05D4 05D6 05DE 05DC 05FC 05FE 05F6 05F4 0574 0576 057E 057C 055C 055E 0556 0554
0D44 0D46 0D4E 0D4C 0D6C 0D6E 0D66 0D64 0DE4 0DE6 0DEE 0DEC 0DCC 0DCE 0DC6 0DC4 0FC4 0FC6 0FCE 0FCC 0FEC 0FEE 0FE6 0FE4 0F64 0F66 0F6E 0F6C 0F4C 0F4E 0F46 0F44 0744 0746 074E 074C 076C 076E 0766 0764 07E4 07E6 07EE 07EC 07CC 07CE 07C6 07C4 05C4 05C6 05CE 05CC 05EC 05EE 05E6 05E4 0564 0566 056E 056C 054C 054E 0546 0544
0D45 0D47 0D4F 0D4D 0D6D 0D6F 0D67 0D65 0DE5 0DE7 0DEF 0DED 0DCD 0DCF 0DC7 0DC5 0FC5 0FC7 0FCF 0FCD 0FED 0FEF 0FE7 0FE5 0F65 0F67 0F6F 0F6D 0F4D 0F4F 0F47 0F45 0745 0747 074F 074D 076D 076F 0767 0765 07E5 07E7 07EF 07ED 07CD 07CF 07C7 07C5 05C5 05C7 05CF 05CD 05ED 05EF 05E7 05E5 0565 0567 056F 056D 054D 054F 0547 0545
0D41 0D43 0D4B 0D49 0D69 0D6B 0D63 0D61 0DE1 0DE3 0DEB 0DE9 0DC9 0DCB 0DC3 0DC1 0FC1 0FC3 0FCB 0FC9 0FE9 0FEB 0FE3 0FE1 0F61 0F63 0F6B 0F69 0F49 0F4B 0F43 0F41 0741 0743 074B 0749 0769 076B 0763 0761 07E1 07E3 07EB 07E9 07C9 07CB 07C3 07C1 05C1 05C3 05CB 05C9 05E9 05EB 05E3 05E1 0561 0563 056B 0569 0549 054B 0543 0541
0D40 0D42 0D4A 0D48 0D68 0D6A 0D62 0D60 0DE0 0DE2 0DEA 0DE8 0DC8 0DCA 0DC2 0DC0 0FC0 0FC2 0FCA 0FC8 0FE8 0FEA 0FE2 0FE0 0F60 0F62 0F6A 0F68 0F48 0F4A 0F42 0F40 0740 0742 074A 0748 0768 076A 0762 0760 07E0 07E2 07EA 07E8 07C8 07CA 07C2 07C0 05C0 05C2 05CA 05C8 05E8 05EA 05E2 05E0 0560 0562 056A 0568 0548 054A 0542 0540
0C40 0C42 0C4A 0C48 0C68 0C6A 0C62 0C60 0CE0 0CE2 0CEA 0CE8 0CC8 0CCA 0CC2 0CC0 0EC0 0EC2 0ECA 0EC8 0EE8 0EEA 0EE2 0EE0 0E60 0E62 0E6A 0E68 0E48 0E4A 0E42 0E40 0640 0642 064A 0648 0668 066A 0662 0660 06E0 06E2 06EA 06E8 06C8 06CA 06C2 06C0 04C0 04C2 04CA 04C8 04E8 04EA 04E2 04E0 0460 0462 046A 0468 0448 044A 0442 0440
0C41 0C43 0C4B 0C49 0C69 0C6B 0C63 0C61 0CE1 0CE3 0CEB 0CE9 0CC9 0CCB 0CC3 0CC1 0EC1 0EC3 0ECB 0EC9 0EE9 0EEB 0EE3 0EE1 0E61 0E63 0E6B 0E69 0E49 0E4B 0E43 0E41 0641 0643 064B 0649 0669 066B 0663 0661 06E1 06E3 06EB 06E9 06C9 06CB 06C3 06C1 04C1 04C3 04CB 04C9 04E9 04EB 04E3 04E1 0461 0463 046B 0469 0449 044B 0443 0441
0C45 0C47 0C4F 0C4D 0C6D 0C6F 0C67 0C65 0CE5 0CE7 0CEF 0CED 0CCD 0CCF 0CC7 0CC5 0EC5 0EC7 0ECF 0ECD 0EED 0EEF 0EE7 0EE5 0E65 0E67 0E6F 0E6D 0E4D 0E4F 0E47 0E45 0645 0647 064F 064D 066D 066F 0667 0665 06E5 06E7 06EF 06ED 06CD 06CF 06C7 06C5 04C5 04C7 04CF 04CD 04ED 04EF 04E7 04E5 0465 0467 046F 046D 044D 044F 0447 0445
0C44 0C46 0C4E 0C4C 0C6C 0C6E 0C66 0C64 0CE4 0CE6 0CEE 0CEC 0CCC 0CCE 0CC6 0CC4 0EC4 0EC6 0ECE 0ECC 0EEC 0EEE 0EE6 0EE4 0E64 0E66 0E6E 0E6C 0E4C 0E4E 0E46 0E44 0644 0646 064E 064C 066C 066E 0666 0664 06E4 06E6 06EE 06EC 06CC 06CE 06C6 06C4 04C4 04C6 04CE 04CC 04EC 04EE 04E6 04E4 0464 0466 046E 046C 044C 044E 0446 0444
0C54 0C56 0C5E 0C5C 0C7C 0C7E 0C76 0C74 0CF4 0CF6 0CFE 0CFC 0CDC 0CDE 0CD6 0CD4 0ED4 0ED6 0EDE 0EDC 0EFC 0EFE 0EF6 0EF4 0E74 0E76 0E7E 0E7C 0E5C 0E5E 0E56 0E54 0654 0656 065E 065C 067C 067E 0676 0674 06F4 06F6 06FE 06FC 06DC 06DE 06D6 06D4 04D4 04D6 04DE 04DC 04FC 04FE 04F6 04F4 0474 0476 047E 047C 045C 045E 0456 0454
0C55 0C57 0C5F 0C5D 0C7D 0C7F 0C77 0C75 0CF5 0CF7 0CFF 0CFD 0CDD 0CDF 0CD7 0CD5 0ED5 0ED7 0EDF 0EDD 0EFD 0EFF 0EF7 0EF5 0E75 0E77 0E7F 0E7D 0E5D 0E5F 0E57 0E55 0655 0657 065F 065D 067D 067F 0677 0675 06F5 06F7 06FF 06FD 06DD 06DF 06D7 06D5 04D5 04D7 04DF 04DD 04FD 04FF 04F7 04F5 0475 0477 047F 047D 045D 045F 0457 0455
0C51 0C53 0C5B 0C59 0C79 0C7B 0C73 0C71 0CF1 0CF3 0CFB 0CF9 0CD9 0CDB 0CD3 0CD1 0ED1 0ED3 0EDB 0ED9 0EF9 0EFB 0EF3 0EF1 0E71 0E73 0E7B 0E79 0E59 0E5B 0E53 0E51 0651 0653 065B 0659 0679 067B 0673 0671 06F1 06F3 06FB 06F9 06D9 06DB 06D3 06D1 04D1 04D3 04DB 04D9 04F9 04FB 04F3 04F1 0471 0473 047B 0479 0459 045B 0453 0451
0C50 0C52 0C5A 0C58 0C78 0C7A 0C72 0C70 0CF0 0CF2 0CFA 0CF8 0CD8 0CDA 0CD2 0CD0 0ED0 0ED2 0EDA 0ED8 0EF8 0EFA 0EF2 0EF0 0E70 0E72 0E7A 0E78 0E58 0E5A 0E52 0E50 0650 0652 065A 0658 0678 067A 0672 0670 06F0 06F2 06FA 06F8 06D8 06DA 06D2 06D0 04D0 04D2 04DA 04D8 04F8 04FA 04F2 04F0 0470 0472 047A 0478 0458 045A 0452 0450
0C10 0C12 0C1A 0C18 0C38 0C3A 0C32 0C30 0CB0 0CB2 0CBA 0CB8 0C98 0C9A 0C92 0C90 0E90 0E92 0E9A 0E98 0EB8 0EBA 0EB2 0EB0 0E30 0E32 0E3A 0E38 0E18 0E1A 0E12 0E10 0610 0612 061A 0618 0638 063A 0632 0630 06B0 06B2 06BA 06B8 0698 069A 0692 0690 0490 0492 049A 0498 04B8 04BA 04B2 04B0 0430 0432 043A 0438 0418 041A 0412 0410
0C11 0C13 0C1B 0C19 0C39 0C3B 0C33 0C31 0CB1 0CB3 0CBB 0CB9 0C99 0C9B 0C93 0C91 0E91 0E93 0E9B 0E99 0EB9 0EBB 0EB3 0EB1 0E31 0E33 0E3B 0E39 0E19 0E1B 0E13 0E11 0611 0613 061B 0619 0639 063B 0633 0631 06B1 06B3 06BB 06B9 0699 069B 0693 0691 0491 0493 049B 0499 04B9 04BB 04B3 04B1 0431 0433 043B 0439 0419 041B 0413 0411
0C15 0C17 0C1F 0C1D 0C3D 0C3F 0C37 0C35 0CB5 0CB7 0CBF 0CBD 0C9D 0C9F 0C97 0C95 0E95 0E97 0E9F 0E9D 0EBD 0EBF 0EB7 0EB5 0E35 0E37 0E3F 0E3D 0E1D 0E1F 0E17 0E15 0615 0617 061F 061D 063D 063F 0637 0635 06B5 06B7 06BF 06BD 069D 069F 0697 0695 0495 0497 049F 049D 04BD 04BF 04B7 04B5 0435 0437 043F 043D 041D 041F 0417 0415
0C14 0C16 0C1E 0C1C 0C3C 0C3E 0C36 0C34 0CB4 0CB6 0CBE 0CBC 0C9C 0C9E 0C96 0C94 0E94 0E96 0E9E 0E9C 0EBC 0EBE 0EB6 0EB4 0E34 0E36 0E3E 0E3C 0E1C 0E1E 0E16 0E14 0614 0616 061E 061C 063C 063E 0636 0634 06B4 06B6 06BE 06BC 069C 069E 0696 0694 0494 0496 049E 049C 04BC 04BE 04B6 04B4 0434 0436 043E 043C 041C 041E 0416 0414
0C04 0C06 0C0E 0C0C 0C2C 0C2E 0C26 0C24 0CA4 0CA6 0CAE 0CAC 0C8C 0C8E 0C86 0C84 0E84 0E86 0E8E 0E8C 0EAC 0EAE 0EA6 0EA4 0E24 0E26 0E2E 0E2C 0E0C 0E0E 0E06 0E04 0604 0606 060E 060C 062C 062E 0626 0624 06A4 06A6 06AE 06AC 068C 068E 0686 0684 0484 0486 048E 048C 04AC 04AE 04A6 04A4 0424 0426 042E 042C 040C 040E 0406 0404
0C05 0C07 0C0F 0C0D 0C2D 0C2F 0C27 0C25 0CA5 0CA7 0CAF 0CAD 0C8D 0C8F 0C87 0C85 0E85 0E87 0E8F 0E8D 0EAD 0EAF 0EA7 0EA5 0E25 0E27 0E2F 0E2D 0E0D 0E0F 0E07 0E05 0605 0607 060F 060D 062D 062F 0627 0625 06A5 06A7 06AF 06AD 068D 068F 0687 0685 0485 0487 048F 048D 04AD 04AF 04A7 04A5 0425 0427 042F 042D 040D 040F 0407 0405
0C01 0C03 0C0B 0C09 0C29 0C2B 0C23 0C21 0CA1 0CA3 0CAB 0CA9 0C89 0C8B 0C83 0C81 0E81 0E83 0E8B 0E89 0EA9 0EAB 0EA3 0EA1 0E21 0E23 0E2B 0E29 0E09 0E0B 0E03 0E01 0601 0603 060B 0609 0629 062B 0623 0621 06A1 06A3 06AB 06A9 0689 068B 0683 0681 0481 0483 048B 0489 04A9 04AB 04A3 04A1 0421 0423 042B 0429 0409 040B 0403 0401
0C00 0C02 0C0A 0C08 0C28 0C2A 0C22 0C20 0CA0 0CA2 0CAA 0CA8 0C88 0C8A 0C82 0C80 0E80 0E82 0E8A 0E88 0EA8 0EAA 0EA2 0EA0 0E20 0E22 0E2A 0E28 0E08 0E0A 0E02 0E00 0600 0602 060A 0608 0628 062A 0622 0620 06A0 06A2 06AA 06A8 0688 068A 0682 0680 0480 0482 048A 0488 04A8 04AA 04A2 04A0 0420 0422 042A 0428 0408 040A 0402 0400
11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63 65 67 69 71 73 75 77 79 81 83 85 87 89 91 93 95 97 99 101 103 105 107 109 111 113 115 117 119 121 123 125 127
10/29/13
CableLabs
195
CM-SP-PHYv3.1-I01-131029
A.2
QAM Constellation Scaling
The CM MUST scale real and imaginary axes of the constellations by the scaling factors given in column 3 of the
table below, to ensure that the mean square value of all QAM constellations are equal to 1.0.
The CMTS MUST scale real and imaginary axes of the constellations by the scaling factors given in column 3 of the
table below, to ensure that the mean square value of all QAM constellations are equal to 1.0.
Table A1 - QAM Constellation Scaling Factors
196
QAM
Constellation
m
Number of bits
BPSK
QPSK
1
2
8-QAM
16-QAM
32-QAM
64-QAM
128-QAM
256-QAM
512-QAM
1024-QAM
10
2048-QAM
11
4096-QAM
12
8192-QAM
13
16384-QAM
14
CableLabs
Scaling
Factor
1
1/2
1/10
1/10
1/20
1/42
1/82
1/170
1/330
1/682
1/1322
1/2730
1/5290
1/10922
10/29/13
Annex B
CM-SP-PHYv3.1-I01-131029
RFoG Operating Mode
The CMTS MUST support the ability to limit the number of simultaneous US transmitters to a single transmitter at a
time.
10/29/13
CableLabs
197
CM-SP-PHYv3.1-I01-131029
Annex C
Additions and Modifications for European Specification

with SC-QAM operation
This section applies to cases where a DOCSIS 3.1 CM or CMTS is operating with Single Carrier QAM (SC-QAM)
operation only, with no OFDM operation. As such, it represents backward compatibility requirements when
operating with DOCSIS 3.0 systems or with the DOCSIS 3.1 PHY disabled. It also applies only to the second
technology option referred to in Section 1.1; for the first option refer to Section 6, and for the third option refer to
Annex D.
As the requirements for a DOCSIS 3.1 CM and CMTS are largely unchanged relative to DOCSIS 3.0 devices for
SC-QAM operation, the requirements for operating with this technology option and in this mode are addressed via
reference to the PHYv3.0 and DRFI specifications, with the exception that the minimum requirement for upstream
and downstream channels has been changed for DOCSIS 3.1 devices.
A DOCSIS 3.1 CM MUST support the CM requirements in Annex B of [DOCSIS PHYv3.0], with the exception
that the minimum requirement for upstream channels is 8, and the minimum requirement for downstream channels
is 24.
A DOCSIS 3.1 CMTS MUST support the CMTS requirements in Annex B of [DOCSIS PHYv3.0] with the
exception that the minimum requirement for upstream channels is 8. A DOCSIS 3.1 CMTS MUST support the
CMTS requirements in Annex B of [DOCSIS DRFI], with the addition that the minimum requirement for
downstream channels is 24.
198
CableLabs
10/29/13
Annex D
CM-SP-PHYv3.1-I01-131029
Additions and Modifications for Chinese Specification with

SC-QAM operation
This annex will be added in a subsequent revision of this specification.
10/29/13
CableLabs
199
CM-SP-PHYv3.1-I01-131029
Annex E
24-bit Cyclic redundancy check (CRC) Code
This section contains a 24-bits CRC code encoding, which is used for NCPs as specified in Section 7.5.14 and Initial
Ranging as specified in Section 7.4.15.1.
The CRC encoder generates the 24 bits parity bits denoted by p0 , p1 , p2 , p3 ,..., p23 for the input bit stream
b0 , b1 ,..., bk 1 .using the following generator polynomial:
g CRC 24 ( x) = x 24 + x 22 + x 20 + x19 + x18 + x16 + x14 + x13 + x11 + x10 + x 8 + x 7 + x 6 + x 3 + x + 1

(127266713 in octal representation) [1], which means in GF(2) the following equation holds
b0 x k +23 + b1 x k +22 + ... + bk 1 x 24 + p0 x 23 + p1 x 22 + ... + p22 x1 + p23 = 0 mod g CRC 24 ( x) .

This 24-bit CRC polynomial is optimized by G. Castangnoli, S. Bruer and M. Hermann in [CMB1993].
200
CableLabs
10/29/13
Appendix I
CM-SP-PHYv3.1-I01-131029
Downstream Frequency Interleaver Sample C Code
In the downstream frequency interleaver C code given below, X is the input data array, Y is the output data array,
and N is the size of each array. It has been written to illustrate each operation clearly, and as such it may not
necessarily be the most efficient implementation.
void Docsis_3_1_Freqeuncy_Interleaver ( int *X, int *Y, int N )
{
int store[128][64];
int i, j, K, k1, base;
int RowCount, RowCount_BR;
int ColumnCount, ColumnCount_BR;
int Last_Column_Size;
int column_rotate[64], rotated_column[128];
// Number of rows is 128
// Number of columns is K
K = (int) (ceil (((double) N)/128.0));
Last_Column_Size = N - ((K-1)*128);
// Generate the column rotation array using the 6-bit LFSR
column_rotate[0]=17;
for (i=0; i<63; i++)
{
int lsb;
lsb = column_rotate[i]&1;
column_rotate[i+1] = (column_rotate[i] >> 1) ^ (lsb << 5) ^ (lsb << 4);
}
base = 0;
// Save data in 2-D store with rows addressed in bit-reversed order
for (RowCount=0; RowCount< 128; RowCount++)
{
RowCount_BR = 0;
for (j=0; j<7; j++)
{
RowCount_BR = RowCount_BR << 1;;
if (((RowCount >> j) & 1) == 1)
RowCount_BR++;
}
if (RowCount_BR < Last_Column_Size)
{
for (j=0; j<K; j++)
store[RowCount_BR][j] = X[base+j];
base += K;
}
else
{
for (j=0; j<K-1; j++)
store[RowCount_BR][j] = X[base+j];
base += (K-1);
}
}
// Rotate columns 0 to K-2
10/29/13
CableLabs
201
CM-SP-PHYv3.1-I01-131029
// Last column that could be partially filled is not rotated

for (j=0; j<(K-1); j++)
{
for (i=0; i<128; i++)
rotated_column[(i + column_rotate[j]) & 0x7F] = store[i][j];
for (i=0; i<128; i++)
store[i][j] = rotated_column[i];
}
// Determine the number of bits k1 in ColumnCount
k1 = 0;
while ((1 << k1) < K)
k1++;
// Address columns in bit-reversed order, but only if bit-reversed number is within range
// Otherwise use non-bit-reversed version
base = 0;
for (ColumnCount=0; ColumnCount<K; ColumnCount++)
{
int ReadColumn;
// Bit Reverse ColumnCount
ColumnCount_BR = 0;
for (i=0; i<k1; i++)
{
ColumnCount_BR = ColumnCount_BR << 1;
if (((ColumnCount >> i) & 1) == 1)
ColumnCount_BR++;
}
if (ColumnCount_BR > (K-1)) // Read from ColumnCount
ReadColumn = ColumnCount;
else
ReadColumn = ColumnCount_BR;
if (ReadColumn == (K-1)) // Last column could be a partially-filled column
{
for (i=0; i<Last_Column_Size; i++)
Y[base+i] = store[i][ReadColumn];
base = base+Last_Column_Size;
}
else
{
for (i=0; i<128; i++)
Y[base+i] = store[i][ReadColumn];
base = base+128;
}
}
return;
}
202
CableLabs
10/29/13
Appendix II
CM-SP-PHYv3.1-I01-131029
Use Cases: Maximum Number of Simultaneous Transmitters
This appendix will be added in a subsequent revision of this specification.
10/29/13
CableLabs
203
CM-SP-PHYv3.1-I01-131029
Appendix III Upstream Time and Frequency Interleaver Sample C Code

The algorithm for generating the sequence of addresses (, f) is described using the following C code segment:
Count_t = 0;
Count_f = 0;
Count_diagonal = 0;
for (idx_t=0;
idx_t<K;
idx_t++)
{
for (idx_f=0;
idx_f<L;
idx_f++)
{
Address.t = Bit_Reverse_Count(&Count_t, K);
Address.f = Bit_Reverse_Count(&Count_f, L);
Count_t = (Count_t + 1) % K1;
Count_f = (Count_f + 1) % L1;
}
Count_diagonal = (Count_diagonal + 1) % K1;
d = Bit_Reverse_Count(&Count_diagonal, K);
Count_t = Count_diagonal;
Count_f = 0;
}
204
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Appendix IV FEC Codeword Selection Algorithm

If the CMTS scheduler wishes to grant a certain number of information bits, it needs to perform the opposite
(forward) calculation in order to determine how many codewords of what sizes are necessary to hold the desired
number of information bits. This is part of the process of determining the grant size. For informative purposes, the
script forward_calc.m is provided to show how the CMTS could perform this calculation. The variable finfo_size is
the input to this script.
% forward_calc script
%
%
%
%
%
set values for codeword sizes

total bits = size including parity
info bits = information bits only
thresholds - if more bits than threshold, shorten this cw instead of
using a smaller one
% short codeword
SHORT_TOTAL_BITS = 1120;
SHORT_INFO_BITS = 840;
SHORT_PARITY_BITS = SHORT_TOTAL_BITS - SHORT_INFO_BITS;
SHORT_TOTAL_THRESH_BITS = SHORT_PARITY_BITS + 1;
SHORT_MIN_INFO_BITS = SHORT_INFO_BITS / 2;
% medium codeword
MED_TOTAL_BITS = 5940;
MED_INFO_BITS = 5040;
MED_PARITY_BITS = MED_TOTAL_BITS - MED_INFO_BITS;
MED_INFO_THRESH_BITS = 2521;
% long codeword
LONG_TOTAL_BITS = 16200;
LONG_INFO_BITS = 14400;
LONG_PARITY_BITS = LONG_TOTAL_BITS - LONG_INFO_BITS;
LONG_TOTAL_THRESH_BITS = 11881;
LONG_INFO_THRESH_BITS = 10081;
% variable finfo_size is input

% set finfo_size to desired input value in workspace
% initialize output variables
flong_cws = 0;
fshortened_long_cws = 0;
fmed_cws = 0;
fshortened_med_cws = 0;
fshort_cws = 0;
fshortened_short_cws = 0;
fother_shortened_cw_bits = 0;
fshortened_cw_bits = 0;
mac_padding = 0;
% intermediate variable to track type of last full codeword
flast_full_cw = '';
% now begin calculation
bits_remaining = finfo_size;
10/29/13
CableLabs
205
CM-SP-PHYv3.1-I01-131029
% if there are no bits at all, we don't want to give any grant - just let
% everything fall through to zero codewords.
%
%
%
%
However, if we have a nonzero number of bits that is not enough to make

a min-size short codeword, we do want to give a grant but we must
allow space for the CM to make a segment that will fill the min-size
short codeword.
if (bits_remaining > 0 && bits_remaining < SHORT_MIN_INFO_BITS)

mac_padding = SHORT_MIN_INFO_BITS - bits_remaining;
bits_remaining = SHORT_MIN_INFO_BITS;
end
% make as many long cws as possible
while bits_remaining >= LONG_INFO_BITS
flong_cws = flong_cws + 1;
bits_remaining = bits_remaining - LONG_INFO_BITS;
flast_full_cw = 'Long';
end
% if remaining bits can make a shortened long cw, do so
if bits_remaining >= LONG_INFO_THRESH_BITS
fshortened_long_cws = 1;
fshortened_cw_bits = bits_remaining;
bits_remaining = 0;
end
% now make as many medium cws as possible
while bits_remaining >= MED_INFO_BITS
fmed_cws = fmed_cws + 1;
bits_remaining = bits_remaining - MED_INFO_BITS;
flast_full_cw = 'Medium';
end
% if remaining bits can make a shortened med cw, do so
if bits_remaining >= MED_INFO_THRESH_BITS
fshortened_med_cws = 1;
bits_remaining = 0;
end
% now make as many short cws as possible
while bits_remaining >= SHORT_INFO_BITS
fshort_cws = fshort_cws + 1;
bits_remaining = bits_remaining - SHORT_INFO_BITS;
flast_full_cw = 'Short';
end
% if there are any bits left, finish with a shortened short cw
if bits_remaining >= 1
fshortened_short_cws = 1;
%
%
%
%
%
we need at least SHORT_MIN_INFO_BITS in a shortened short

codeword; if we don't have enough, we will borrow from the
immediately preceding full codeword, which will become another
shortened codeword.
Note that we will always borrow SHORT_MIN_INFO_BITS.
if bits_remaining >= SHORT_MIN_INFO_BITS

% no need to borrow bits
bits_remaining = 0;
206
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
else
% identify type/size of last full codeword
switch flast_full_cw
case 'Long'
% change last full cw to a shortened cw
flong_cws = flong_cws - 1;
fshortened_long_cws = fshortened_long_cws + 1;
% number of bits in that cw is reduced by
% SHORT_MIN_INFO_BITS
fother_shortened_cw_bits = LONG_INFO_BITS - ...
% put those bits plus bits_remaining into the last
% shortened cw
fshortened_cw_bits = SHORT_MIN_INFO_BITS +bits_remaining;
bits_remaining = 0;
case 'Medium'
% same steps as with long
fmed_cws = fmed_cws - 1;
fshortened_med_cws = fshortened_med_cws + 1;
fother_shortened_cw_bits = MED_INFO_BITS - ...
bits_remaining = 0;
case 'Short'
% also same as long
fshort_cws = fshort_cws - 1;
fshortened_short_cws = fshortened_short_cws + 1;
fother_shortened_cw_bits = SHORT_INFO_BITS - ...
bits_remaining = 0;
end
end
end
Tables showing the number and size of codewords to be used for grant sizes from 1 bit up to two full long
codewords are provided in [PHYv3.1_CODECHECK].
10/29/13
CableLabs
207
CM-SP-PHYv3.1-I01-131029
Appendix V
CMTS Proposed Configuration Parameters

Table V1 - CMTS Proposed Configuration Parameters
FFT
Cyclic Prefix
Samples
(Ncp)
192
256
512
4K
768
1024
192
256
512
8K
768
1024
208
Roll-Off Period
Samples
(Nrp)
Bandedge
Exclusion
Sub-band
(MHz)
Lower Edge
Exclusion
Sub-band
(subcarriers)
Upper Edge
Exclusion
Sub-band
(subcarriers)
64
3.650
201
200
128
2.000
168
167
64
3.650
201
200
128
2.000
168
167
192
1.450
157
156
64
3.650
201
200
128
1.950
167
166
192
1.400
156
155
256
1.100
150
149
64
3.600
200
199
128
1.950
167
166
192
1.400
156
155
256
1.100
150
149
64
3.600
200
199
128
1.900
166
165
192
1.350
155
154
256
1.050
149
148
64
3.400
392
391
128
1.750
326
325
64
3.400
392
391
128
1.750
326
325
192
1.225
305
304
64
3.375
391
390
128
1.750
326
325
192
1.200
304
303
256
0.925
293
292
64
3.375
391
390
128
1.750
326
325
192
1.200
304
303
256
0.925
293
292
64
3.350
390
389
128
1.725
325
324
192
1.200
304
303
256
0.925
293
292
CableLabs
10/29/13
CM-SP-PHYv3.1-I01-131029
Appendix VI Acknowledgements
On behalf of the cable industry and our member companies, CableLabs would like to thank the numerous
individuals that contributed to the development of this specification. In particular, we want to extend our sincere
appreciation and gratitude to the following team members for their contributions in developing the DOCSIS 3.1
Physical Layer specifications
Contributor
Company Affiliation
Contributor
Company Affiliation
Brian Kurtz
Altera
Boris Brun
Harmonic
Ayham Al-Banna
Arris
Jim Chen
Huawei
Doc Cummings
Arris
Syed Rahman
Huawei
Mike Emmendorfer
Arris
Guangsheng Wu
Huawei
Bruce Currivan
Broadcom
Bernard Arambepola
Intel
Roger Fish
Broadcom
Satish Mudugere
Intel
Avi Kliger
Broadcom
Shaul Shulman
Intel
Tom Kolze
Broadcom
Noam Tal
Intel
Rich Prodan
Broadcom
Joel Birkeland
Maxim
Alberto Campos
CableLabs
Ajay Kuckreja
Maxim
Belal Hamzeh
CableLabs
Suman Musunuru
Maxim
Volker Leisse
CableLabs
Scott Carichner
Maxlinear
Kevin Luehrs
CableLabs
Curtis Ling
Maxlinear
Joey Padden
CableLabs
Sridhar Ramesh
Maxlinear
Matthew Schmitt
CableLabs
Scott Walley
Maxlinear
Tom Williams
CableLabs
Rob Howald
Motorola
Guy Cohen
Capacicom
Rob Thompson
Motorola
Daniel Wacjer
Capacicom
Stefan Brueck
Qualcomm
Dave Fox
Casa
Salvo Pennisi
Qualcomm
Brett Hansmeier
Casa
Marc Werner
Qualcomm
John Chapman
Cisco
George Hart
Rogers Communications
Ron Hranac
Cisco
Marty Davidson
SCTE
J Hang in
Cisco
David Armstrong
SED Systems
Rick Meller
Cisco
Dan Coode
SED Systems
Bill Wall
Cisco
Martin Lovell
Sony
Tom Staniec
Cohere
Lothar Stadelmeier
Sony
Phil Chang
Comcast
Jean-Noel Rozec
ST
Peter Hutnick
Comcast
Neil Robertson
ST Microelectronics
Saifur Rahman
Comcast
Jim Dietz
Texas Instruments
Jorge Salinger
Comcast
Joel Halbert
Texas Instruments
Joe Solomon
Comcast
Carey Ritchey
TI
Dave Urban
Comcast
Carissa Sipp
TI
Ony Anglade
Cox Communications
Paul Brooks
Time Warner Cable
Eugene Dai
Cox Communications
Kirk Erichsen
Time Warner Cable
Jeff Finkelstein
Cox Communications
Colin Howlett
Vecima
Hank Blauvelt
Emcore
Chris Dick
Xilinx
Adi Bonen
Harmonic
Joe Palermo
Xilinx
10/29/13
CableLabs
209
CM-SP-PHYv3.1-I01-131029
Additionally, Cablelabs would like to thank our member MSOs for their continued support in driving the
specification development and decision making process.
210
CableLabs
10/29/13

CM SP PHYv3.1 I01 131029

Uploaded by

Copyright:

Available Formats

CM SP PHYv3.1 I01 131029

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

CM SP PHYv3.1 I01 131029

Uploaded by

Copyright:

Available Formats

Data-Over-Cable Service Interface Specifications

Physical Layer Specification

CableLabs . Use and understanding of this document may require

Data-Over-Cable Service Interface Specifications

Physical Layer Specification

Document Status Sheet

Key to Document Status Codes

An incomplete document, designed to guide discussion and generate feedback

A document in specification format considered largely complete, but lacking

A generally public document that has undergone Member and Technology

A static document, reviewed, tested, validated, and closed to further engineering

Data-Over-Cable Service Interface Specifications

NORMATIVE REFERENCES .....................................................................................................................................7

TERMS AND DEFINITIONS............................................................................................................................... 9

ABBREVIATIONS AND ACRONYMS ............................................................................................................... 21

OVERVIEW AND FUNCTIONAL ASSUMPTIONS .............................................................................................. 26

PHY SUBLAYER FOR SC-QAM ........................................................................................................................ 31

Physical Layer Specification

6.2.17 Spectral Shaping ..................................................................................................................................32

PHY SUBLAYER FOR OFDM ........................................................................................................................... 35

Data-Over-Cable Service Interface Specifications

7.4.15 Ranging ................................................................................................................................................76

PHYMAC CONVERGENCE .......................................................................................................................... 166

PROACTIVE NETWORK MAINTENANCE....................................................................................................... 179

Physical Layer Specification

QAM CONSTELLATION MAPPINGS .................................................................................................. 186

RFOG OPERATING MODE ................................................................................................................ 197

24-BIT CYCLIC REDUNDANCY CHECK (CRC) CODE ............................................................................ 200

Data-Over-Cable Service Interface Specifications

FIGURE 737 - PILOT PATTERN FOR SUBSLOTS WITH 16 SUBCARRIERS .......................................................................................92

Physical Layer Specification

FIGURE A7 - 128-QAM CONSTELLATION MAPPING OF {} ..................................................................188

Data-Over-Cable Service Interface Specifications

Physical Layer Specification

Introduction and Purpose

Broadband Access Network

Data-Over-Cable Service Interface Specifications

Network and System Architecture

Back Office Network

Figure 11 - The DOCSIS Network

Certificate Revocation server provides certificate status.

Physical Layer Specification

The Syslog server collects messages pertaining to the operation of devices.

Figure 12 - Transparent IP Traffic Through the Data-Over-Cable System

Data-Over-Cable Service Interface Specifications

NOTE: Lighter shaded areas are

MAC & Upper Layer

Figure 1-3 - Data-Over-Cable Reference Architecture

DOCSIS 3.1 Documents

Physical Layer Specification

Media Access Control and Upper Layer Protocols Interface Specification

Operations Support System Interface-CM Specification

Operations Support System Interface-CMTS Specification

Cable Modem CPE Interface Specification

This specification defines the interface for the physical layer.

Physical Layer Specification

Table 12 - DOCSIS 3.1 Related Specifications

Downstream Radio Frequency Interface Specification

DOCSIS Timing Interface Specification