5G RAN NR Air Interface
5G RAN NR Air Interface
Objectives of Chapter 4
After this chapter the participants will be able to:
4 Detail the physical procedures for user data transmissions
4.1 Detail the reference signals related to user plane transmissions (DMRS, CSI-RS, TRS, PTRS, SRS)
4.2 Explain Type A and Type B transmissions
4.3 Explain HARQ codebook principles and Code Block Group (CBG) based retransmissions
4.4 Describe UL and DL scheduling principles and resource allocation in frequency and time domains
4.5 Explain link adaptation principles
4.6 Describe LDPC channel coding
Logical Channels
Priority handling, payload
selection MAC
Mux/concatenation Demultiplexing
Retransmission control
Hybrid
(Fast) ARQ
Hybrid ARQ Hybrid
(Fast) ARQ
Hybrid ARQ
MAC scheduler
Redundancy
version
Transport Channels
PHY PHY
TBS Rate matching Coding
Coding+ RM CodingDecoding
(on-the-fly) + RM
Modulation scheme
Data modulation
Modulation
Data modulation
Demodulation
Antenna and resource
assignment Antenna and resrouce mapping Antenna and resrouce mapping
Antenna and resource mapping Antenna and resource
demapping
Physical Channels
gNodeB UE
© Ericsson AB 2018 | | LZU1082750 PA1 | Figure 3-3
NR Channel Structure
DL UL
MIB SIBs
MAC
Transport
Channels
PCH BCH DL-SCH UL-SCH RACH
DCI UCI
Physical
Channels
PBCH PDSCH PDCCH PUSCH PUCCH PRACH
Reference and
Sync Signals
PSS SSS DMRS PTRS DMRS DMRS CSI-RS /TRS SRS PTRS DMRS DMRS PRACH
Preamble
SSB
DL Power Control
Number of DM- DM-RS configuration DM-RS epre- The number of PDSCH layers
RS CDM groups type 1 configuration type Ratio 1 2 3 4 5 6
without data 2
0 0 3 4.77 6 7 7.78
1 0 dB 0 dB
2 -3 dB -3 dB
1 0 0 0 0 0 0
3 - -4.77 dB 2 reserved
3 reserved
© Ericsson AB 2018 | | LZU1082750 PA1 | Figure 3-5
38.214-f20:
The gNodeB determines the downlink transmit EPRE.
For the purpose of SS-RSRP, SS-RSRQ and SS-SINR measurements, the UE may
assume downlink EPRE is constant across the bandwidth. For the purpose of SS-
RSRP, SS-RSRQ and SS-SINR measurements, the UE may assume downlink EPRE is
constant over SSS carried in different SS/PBCH blocks. For the purpose of SS-RSRP,
SS-RSRQ and SS-SINR measurements, the UE may assume that the ratio of SSS
EPRE to PBCH DM-RS EPRE is 0 dB.
For the purpose of CSI-RSRP, CSI-RSRQ and CSI-SINR measurements, the UE may
assume downlink EPRE of a port of CSI-RS resource configuration is constant across
the configured downlink bandwidth and constant across all configured OFDM
symbols.
The downlink SS/PBCH SSS EPRE can be derived from the SS/PBCH downlink
transmit power given by the parameter SS-PBCH-BlockPower provided by higher
layers. The downlink SSS transmit power is defined as the linear average over the
power contributions (in [W]) of all resource elements that carry the SSS within the
operating system bandwidth.
When the UE is scheduled with PT-RS ports associated with the PDSCH and when
the PT-RS port is associated to nPTRS,DMRS DM-RS ports,
- if the UE is configured with the higher layer parameter epre-Ratio, the ratio of PT-
RS EPRE to PDSCH EPRE per layer per RE for PT-RS port (pPTRS ) is given by
Table 4.1-2 according to the epre-Ratio, the PT-RS scaling factor BetaPTRS
specified in subclause 7.4.1.2.2 of [4, TS 38.211] is given by BetaPTRS = 10^(-
pPTRS/20).
- otherwise, the UE shall assume epre-Ratio is set to state '0' in Table
4.1-2 if not configured.
Resource Allocation
— Resource allocation type 0 – bitmap (RIV), each bit corresponds to a group of RBs (RBGs)
— Resource allocation type 1 – start and length of RB allocation (SLIV)
— The type to use is RRC configured (always 0, always 1, dynamic selection of 0/1)
— Uplink transmissions limited to contiguous allocations in Rel-15
1 0 0 1 1 1 0 1 0 0 0 1 0
Type 0
Bitmap
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Type 1
RIV: Resource Indication Value, start and length of contiguously allocated RBs
© Ericsson AB 2018 | | LZU1082750 PA1 | Figure 3-8
— UL Type 0
— Only for CP-OFDM (Transform precoding disabled) Carrier Bandwidth
Part Size
Nominal RBG size
Configuration 1
Nominal RBG size
Configuration 2
— UL Type 1
— Both DFT-S-OFDM and CP-OFDM
— RIV indicates start and length of contiguous RBs
— Configured grant
— Higher layers determine the allocation in ConfiguredGrantConfig (PUSCH Type 1) or MAC config
(PUSCH Type 2)
38.214-f20:
Resource allocation in frequency domain
Two downlink resource allocation schemes, type 0 and type 1, are supported. The UE
shall assume that when the scheduling grant is received with DCI format 1_0, then
downlink resource allocation type 1 is used.
If the scheduling DCI is configured to indicate the downlink resource allocation type
as part of the Frequency domain resource assignment field by setting a higher layer
parameter resourceAllocation in pdsch-Config to 'dynamicswitch', the UE shall use
downlink resource allocation type 0 or type 1 as defined by this DCI field. Otherwise
the UE shall use the downlink frequency resource allocation type as defined by the
higher layer parameter resourceAllocation.
For a PDSCH scheduled with a DCI format 1_0 in any type of PDCCH common search
space, regardless of which bandwidth part is the active bandwidth part, RB
numbering starts from the lowest RB of the CORESET in which the DCI was received.
For a PDSCH scheduled otherwise, if a bandwidth part indicator field is not
configured in the scheduling DCI, the RB indexing for downlink type 0 and type 1
resource allocation is determined within the UE's active bandwidth part. If a
bandwidth part indicator field is configured in the scheduling DCI, the RB indexing for
downlink type 0 and type 1 resource allocation is determined within the UE's
bandwidth part indicated by bandwidth part indicator field value in the DCI. The UE
shall upon detection of PDCCH intended for the UE determine first the downlink
carrier bandwidth part and then the resource allocation within the bandwidth part.
Time-domain allocation
— Specification structure supports ‘any’ combination of start, length, and mapping type
if (L 1) 7 then
SLIV 14 (L 1) S
where 0 L 14 S
© Ericsson AB 2018 | | LZU1082750 PA1 | Figure 3-13
RRC configures a table of K1 values, and then an index into this table is transmitted
in the DCI.
2 - - Default B
(needed before configuration) SI-RNTI Type0A common
3
1
-
No
-
-
Default C
Default A
2 No - Default B
— Or RRC TimeDomainAllocationList in either 3
1,2,3
No
Yes
-
-
Default C
pdsch-
Table defines which PDSCH time domain resource allocation configuration to apply.
Either a default PDSCH time domain allocation A, B or C according to tables on
following slides is applied, or the higher layer configured pdsch-
TimeDomainAllocationList in either pdsch-ConfigCommon or pdsch-Config is applied.
Time domain Allocation A (normal CP) Time domain Allocation B Time domain Allocation C
Row dmrs- PDSCH K0 S L Row dmrs- PDSCH K0 S L Row dmrs- PDSCH K0 S L
index TypeA- mapping index TypeA- mapping index TypeA- mapping
Position type Position type Position type
1 2 Type A 0 2 12 1 2,3 Type B 0 2 2 1 (Note 2,3 Type B 0 2 2
3 Type A 0 3 11 2 2,3 Type B 0 4 2 1)
2 2 Type A 0 2 10 3 2,3 Type B 0 6 2 2 2,3 Type B 0 4 2
3 Type A 0 3 9 4 2,3 Type B 0 8 2 3 2,3 Type B 0 6 2
3 2 Type A 0 2 9 5 2,3 Type B 0 10 2 4 2,3 Type B 0 8 2
3 Type A 0 3 8 6 2,3 Type B 1 2 2 5 2,3 Type B 0 10 2
4 2 Type A 0 2 7 7 2,3 Type B 1 4 2 6 Reserved
3 Type A 0 3 6 8 2,3 Type B 0 2 4 7 Reserved
5 2 Type A 0 2 5 9 2,3 Type B 0 4 4 8 2,3 Type B 0 2 4
3 Type A 0 3 4 10 2,3 Type B 0 6 4 9 2,3 Type B 0 4 4
6 2 Type B 0 9 4 11 2,3 Type B 0 8 4 10 2,3 Type B 0 6 4
3 Type B 0 10 4 12 2,3 Type B 0 10 4 11 2,3 Type B 0 8 4
7 2 Type B 0 4 4 (Note 12 2,3 Type B 0 10 4
3 Type B 0 6 4 1) 13 (Note 2,3 Type B 0 2 7
8 2,3 Type B 0 5 7 13 2,3 Type B 0 2 7 1)
9 2,3 Type B 0 5 2 (Note 14 (Note 2 Type A 0 2 12
10 2,3 Type B 0 9 2 1) 1) 3 Type A 0 3 11
11 2,3 Type B 0 12 2 14 2 Type A 0 2 12 15 (Note 2,3 Type A 0 0 6
12 2,3 Type A 0 1 13 (Note 3 Type A 0 3 11 1)
13 2,3 Type A 0 1 6 1) 16 (Note 2,3 Type A 0 2 6
14 2,3 Type A 0 2 4 15 2,3 Type B 1 2 4 1)
15 2,3 Type B 0 4 7 16 Reserved Note 1: The UE may assume that this PDSCH resource allocation is
16 2,3 Type B 0 8 4 Note 1: If the PDSCH was scheduled with SI-RNTI in not used, if the PDSCH was scheduled with SI-RNTI in PDCCH
PDCCH Type0 common search space, the UE may assume Type0 common search space
that this PDSCH resource allocation is not applied
— RRC (pusch-TimeDomainResourceAllocationList)
— K2 : DCI slot -> PUSCH slot
— SLIV : Start symbol (S) and Length (L, number of consecutive symbols)
K2
PUSCH
slot n slot n+1
…
DCI
if (L 1) 7 then
SLIV 14 (L 1) S
else
where 0 L 14 S
© Ericsson AB 2018 | | LZU1082750 PA1 | Figure 3-16
HARQ Principle
› NR supports adaptive and asynchronous HARQ in
both UL and DL Earliest possible
re-transmission
› The number of HARQ processes is configured to a UE DL DCI DCI
with at most 16 HARQ processes
Fast A/N
› NR defines 2 UE categories (w.r.t. HARQ processing UL
time line), initially most (all?) UEs will support
baseline category
Upon receiving this “Fast HARQ” feedback, the network should act on the received
information by e.g. either – in case of a (probably) unsuccessful decoding –
retransmitting the same data on the same HARQ process or - in case of a (probably)
successful decoding – transmitting new data on another (or possibly the same -in
case no new is available-) HARQ process. The “fast HARQ” feedback is assumed to
be transmitted on a PUSCH or PUCCH resource.
For the DL the maximum number of HARQ processes is RRC configurable. For the UL
it is not configurable:
Max number of DL HARQ processes per carrier signalled in DCI is 16
For downlink, a maximum of 16 HARQ processes per cell is supported by the UE. The
number of processes the UE may assume will at most be used for the downlink is
configured to the UE for each cell separately by higher layer parameter nrofHARQ-
processesForPDSCH, and when no configuration is provided the UE may assume a
default number of 8 processes.
Max number of UL HARQ processes per carrier signalled in DCI is 16
For uplink, a maximum of 16 HARQ processes per cell is supported by the UE
(38.214, section 6.1). There is no possibility to indicate the max number of used
HARQ processes in the uplink direction, in contrast to downlink.
HARQ Codebook
— The number of bits to feed back depends on the number of downlink transmissions…
…and some of these downlink transmission might be missed by the UE
— HARQ Codebook
— Semi-static or dynamic
HARQ Codebook
— Semi-static codebook (Type 1) — Dynamic codebook (Type 2)
17 ACKs
LDPC for NR
CB size - granularity 8, byte aligned 1. For a small TB without CB segmentation, TB CRC is done at L1 and
no CB CRC
code rate – range [1/3, 22/25] for BG1
(supported by PCM) [1/5, 2/3] for BG2 2. Filler bits are skipped during rate matching and not transmitted
© Ericsson AB 2018 | | LZU1082750 PA1 | Figure 3-24
MAC L1
CB
LDPC Encoding process chain (per code block)
Based
TB Code Block I/F + CB CRC LDPC Rate Matching Scrambling
+TB CRC1 Segmentation and Interleaving (CB based)
+ Filler bits2 Encoder
BG1
2/3
1/4
BG2
292 3824
Note: (1) is the transport block size before CRC attachment, determined by scheduler
(2) is indicated by MCS which is determined by scheduler
(3) BG1 has mother code rate of 1/3, repetition is needed to reach lower code rate
© Ericsson AB 2018 | | LZU1082750 PA1 | Figure 3-26 (4) is the effective code rate above which UE can skip decoding
Base graph #2: 42x52, Rmin,2=1/5 R=2/3 R=1/2 R=1/3 R=1/4 R=1/5
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52
1 1 1 1 1 1 1 1 1
2 1 1 1 1 1 1 1 1 1 1
3 1 1 1 1 1 1 1 1
4 1 1 1 1 1 1 1 1 1 1
5 1 1 1 1
6 1 1 1 1 1 1
7 1 1 1 1 1 1
8 1 1 1 1 1 1
9 1 1 1 1
10 1 1 1 1 1
11 1 1 1 1 1
12 1 1 1 1 1
13 1 1 1 1
14 1 1 1 1 1
15 1 1 1 1 1
16 1 1 1 1
17 1 1 1 1 1
18 1 1 1 1 1
19 1 1 1 1
20 1 1 1 1
21 1 1 1 1
22 1 1 1 1
23 1 1 1
24 1 1 1 1
25 1 1 1 1
26 1 1 1
27 1 1 1 1 1
28 1 1 1
29 1 1 1 1
30 1 1 1
31 1 1 1 1 1
32 1 1 1
33 1 1 1 1
34 1 1 1 1
35 1 1 1 1
36 1 1 1 1
37 1 1 1 1
38 1 1 1
39 1 1 1 1
40 1 1 1 1
41 1 1 1 1
42 1 1 1 1
Low latency in NR
Scheduling assignment for
ongoing transmission
— Lower latency
— Faster scheduling (can occur at any time), typically together with shorter duration
Type B scheduling are transmissions with duration substantially shorter than the slot
length, and also UE monitoring PDCCH in many more locations
— URLLC
— Type B scheduling
— Depending on latency requirements PDCCH is
PDSCH (data)
monitored at any OFDM symbol or just a subset
© Ericsson AB 2018 | | LZU1082750 PA1 | Figure 3-31
Link Adaptation
— Link Adaptation (LA) for PDSCH and PUSCH consists of the following:
— Inner loop MCS selector targeting a fixed Block Error Rate (BLER) of 10% for all HARQ
transmissions
— Outer loop channel quality corrector based on HARQ ACK/NACK feedback to enforce the BLER
target
— For PDSCH, CSI reported by UE (PMI, CRI, CQI, RI) is used as input for LA
— For PUSCH, signal power and noise-plus-interference is measured in the gNodeB and used as input for
LA, together with UE-reported PHR.
Transport-block sizes
Note: can limit the possibility for very low code rates
(Table 5.1.3.2-2) 23
24
208
224
53
54
984
1032
83
84
2728
2792
25 240 55 1064 85 2856
— Otherwise, use a formula to determine TBS 26 256 56 1128 86 2976
27 272 57 1160 87 3104
28 288 58 1192 88 3240
29 304 59 1224 89 3368
30 320 60 1256 90 3496
Ninf o 24
Ninf' o max 3840, 2n round n log2 N inf o 24 5
2
n
N ' 24 N ' 24
TBS 8· C· inf o 24 C inf o If R<=1/4
8· C 3816
N ' 24 N ' 24
TBS 8· C· inf o 24 C inf o If Ninfo>8424
8· C 8424
N ' 24
TBS 8· inf o 24 Otherwise
8
DM-RS
— Type A mapping — Type B
— PDSCH Mapping TypeA — PDSCH Mapping TypeB
— Mapping is relative to slot — Mapping is relative to PDSCH/PUSCH start
— Up to 8 (12) ports, depending on DM-RS type — Release 15 defines Type B PDSCH/PUSCH
— A UE is configured with first front-loaded DM- with 2, 4, and 7 symbols
RS in either 3rd or 4th symbol
— In addition it can be configured with 3 2 symbol PxSCH
additional DM-RS symbols
— Placement depends on PDSCH/PUSCH stop 4 symbol PxSCH
1 additional DM-RS symbols
1 or 2 front-loaded can be configured for
DM-RS 7 symbol PxSCH
symbols
symbols
1 or 2 front-loaded
plus late DM-RS 1st/2nd and 5th/6th symbol
symbols
2 or 3 additional DM-
RS for high Doppler symbols
38.211-f20:
For PDSCH mapping type B
- if the PDSCH duration is 2, 4, or 7 OFDM symbols for normal cyclic
prefix or 2, 4, 6 OFDM symbols for extended cyclic prefix, and the PDSCH allocation
collides with resources reserved for a CORESET, shall be incremented such that the
first DM-RS symbol occurs immediately after the CORESETand
- if the PDSCH duration is 4 symbols, the UE is not expected to receive a
DM-RS symbol beyond the third symbol,
- if the PDSCH duration is 7 symbols for normal cyclic prefix or 6 symbols
for extended cyclic prefix,
- the UE is not expected to receive the first DM-RS beyond the fourth
symbol, and
- if one additional single-symbol DM-RS is configured, the UE only
expects the additional DM-RS to be transmitted on the 5th or 6th symbol when the
front-loaded DM-RS symbol is in the 1st or 2nd symbol, respectively, of the PDSCH
duration, otherwise the UE should expect that the additional DM-RS is not
transmitted.
- if the PDSCH duration is 2 or 4 OFDM symbols, only single-symbol DM-
RS is supported.
10 10
11 11
12 12
13 13
14 14
0 13 0 13
8 8
Mapping type B
9 9
10 10
(PUSCH
11 11
example)
12 12
13 13
14 14
0 13 0 13
DM-RS Sequences
Xm+5 Xm+12
Device 2
xm Xm+7
Device 1
x0 xM-1
-
1004 +
CDM group 2
© Ericsson AB 2018 | | LZU1082750 PA1 | Figure 3-43 1005
in frequency + + +
+
Length-2 OCC in time +
+
+ + +
- - -
+ + +
+ + +
+ + +
- + - - + +
1000 + 1000 + +
+ + + + + +
CDM group 0 - CDM group 0 - -
+ + +
- -
1001
— Single/double symbol
CDM group 1 +- CDM group 1 - - + -
1003 ++ + + -
1003 +
- +
+ -
+ -
- - +
+ + + + + +
+ + +
1000 +
+ 1000 + +
+ +
CDM group 0 - CDM group 0 - +
+ + +
1001 + +
- 1001 - -
+ + + +
+ + +
+ + +
+ + +
1002 1002
- - -
CDM group 1 + CDM group 1 + +
- - -
+ + + + + +
1003 + 1003 + +
- - -
1004 + 1004 + +
+ -
CDM group 2 CDM group 2 + -
1005 1005
- +
+ - + -
+ -
1006 +
+
-
-
CDM group 0 - +
+ -
1007 - +
+ - + -
+ -
+ -
+ -
1008 - +
CDM group 1 + -
- +
+ - + -
1009 + -
- +
+ -
1010
© Ericsson AB 2018 | | LZU1082750 PA1 | Figure 3-44 CDM group 2
1011
Co-scheduled device
No co-scheduled CDM group – reuse for data Co-scheduled CDM groups – leave resource elements empty
subcarriers
— A UE can be configured in DL and UL with PT-RS
to track phase variations (mainly for hi-band) DM-RS
PT-RS
— PT-RS are dense in time
— every, every 2nd or every 4th symbol
— depending on scheduled MCS
— PT-RS are sparse in frequency
— 1 subcarrier every, every 2nd or every 4th PRB
— depending on scheduled BW
— PT-RS not present at low scheduled MCSs and
BWs or for RA, SI and paging messages
symbols
Example: PT-RS occur every 2nd PRB and every symbol
Phase Tracking Reference Signal (PT-RS) is a UE-specific RS which aims for phase
rotation estimation
Uses:
PTRS is used for phase noise effects compensation
Presence:
PTRS presence is indicated by RRC signaling
PTRS is always associated with DMRS and PDSCH
At most two PTRS ports per user (to support multi-panel)
At most six orthogonal PTRS ports for MU-MIMO
Patterns for CP-OFDM:
Time densities: 1, 1/2 and 1/4 (1 PTRS every, every 2nd and every 4th symbol)
Freq. densities: 1/2 and 1/4 (PTRS subcarrier every, every 2nd and 4th PRB)
Time/freq. density is associated with DCI parameters using Tables
Predefined thresholds for tables
UE can suggest by RRC threshold to override the predefined thresholds
UE can suggest different thresholds for DL and UL and per BWP
PT-RS
Scheduled MCS Time density
0<=MCS<MCS1 PTRS OFF
MCS1<=MCS<MCS2 1/4
MCS2<=MCS<MCS3 1/2
— Can be seen as an extension to DM-RS MCS3<=MCS < MCS4 1
Scheduled BW Freq. density
— Can only occur in combination with DM-RS RB<RB0 PTRS OFF
RB0<=RB<RB1 1/2
RB1<=RB 1/4
PT-RS
one slot
Phase Tracking Reference Signal (PT-RS) is a UE-specific RS which aims for phase
rotation estimation.
So, PTRS is used for phase noise compensation in hi-band deployments
0 1 2 3 4 5 6 7 8 9 10 11 12 13
— NR has no CRS to use for fine time-frequency tracking 0
0
1
1
2
2
3
3
4
4
5
5
6
6
7
7
8
8
9
9
10
10
11
11
12
12
13
13
0 1 2 3 4 5 6 7 8 9 10 11 12 13
0 1 2 3 4 5 6 7 8 9 10 11 12 13
0 1 2 3 4 5 6 7 8 9 10 11 12 13
2 adjacent slots
0 1 2 3 4 5 6 7 8 9 10 11 12 13
0 1 2 3 4 5 6 7 8 9 10 11 12 13
0 1 2 3 4 5 6 7 8 9 10 11 12 13
7 8 9 10 11 12 13 0 1 2 3 4 5 6 7 8 9 10 11 12 13 0 1 2 3 4 5 6 7 8 9 10 11 12 13 0 1 2 3 4 5 6
© Ericsson AB 2018 | | LZU1082750 PA1 | Figure 3-48 symbols
TRS basics
Use cases:
Fine Time synchronization
Needed e.g. for channel estimation and demodulation
Rough time synchronization achieved by other means e.g. based on SS-block
The TRS comes in bursts with a periodicity of 10ms, 20ms, 40ms or 80ms
Each burst consists of one or two slots
Each slot in the TRS burst looks the same
The TRS is present in two OFDM symbols in each slot in the TRS burst
There are three alternative OFDM symbol positions within a slot
Symbol indices (4,8), (5,9) or (6,10). Note that the first symbol in a slot
has index 0.
The bandwidth of the TRS is min(50RB,UE BWP) or UE BWP
Every fourth subcarrier is used for the TRS
The subcarrier offset is configurable
Gold sequences are used to ‘randomize’ the TRS and control correlation properties
TRS Format
CSI-RS resource #3
CSI-RS resource #4
CSI-RS resource #1
CSI-RS resource #2
— Transmitted in DL
— Used by UEs to acquire channel state information (CSI)
— Transmitted with e.g. 5, 10, 20, 40 … slots periodicity
— RRC-configured, RRC-activated
CSI period = 10 slots, Offset = 3
— Semi-persistent reporting
— RRC-configured, MAC-CE-activated
— Reports on PUCCH (or ‘re-routed’ to PUSCH)
— Aperiodic reporting
— One shot, triggered by DCI
— Reports on PUSCH
© Ericsson AB 2018 | | LZU1082750 PA1 | Figure 3-54
CSI-RS
— Single-port CSI-RS
One slot
OFDM symbols
— High degree of flexibility 2xCDM + 4xFDM 2xCDM + 2xFDM + 2xTDM 4xCDM + 2xFDM
— Example:
three possibilities for an 8-port CSI-RS
— Freq. density: every or every second RB One slot
AP 0‐1 (2xCDM) AP 2‐3 (2xCDM) AP 0‐1 (2xCDM) AP 2‐3 (2xCDM) AP 0‐3 (4xCDM) AP 4‐7 (4xCDM)
— Zero-power CSI-RS
— Unused resource, the UE can make no assumption on the contents
— In essence a tool for rate matching around certain resource elements
— Configured in same way as NZP-CSI-RS
— CSI-IM
— Used for interference measurements
— Should typically correspond to a ZP-CSI-RS
— Time-domain configured in same way as NZP-CSI-RS
One slot One slot
Mapping to antennas
…
3000 +P/2 3000 + P/2 +1 3000 +P -1
CSI reference signals are transmitted on =1, 2, 4, 8, 12, 16, 24 "or" 32 ports using
e.g. port numbers =3000,3001,…,3000+ −1.
Port numbers are assumed to be mapped to virtual antennas with two different
polarizations using the convention shown in figure, in other words the first half of the
port numbers are mapped to virtual antennas with the first polarization, and the 2nd
half to virtual antennas with the second polarization. Note: Figure shows +45/-45
polarization only as an example. For the purposes of mapping port numbers to the
REs of the aggregated CSI-RS resource, a length-P port number vector is defined.
For the case of >1, the elements _ of this vector are given by
_2 &=3000+
_(2 +1)&=3000+ ⁄2+
&=0,1,…, ⁄2−1
For the case of a single-port only, the port number vector becomes a scalar given by
=3000. With this definition, port numbers corresponding to the same physical cross-
pole are contained within a single CSI-RS unit. For example, for P = 4 ports, the port
number vector is given by =[■8(3000&3002&3001&3003)].
CSI-RS
- Example “beam space”
1 symbol resource 2 symbol resource Beamformed CSI-RS
CSI-RS resource data
TD‐OCC
frequency
FD‐OCC
OFDM symbols
4 symbol resource (adjacent) 4 symbol resource (pairwise adjacent)
time
CSI-RS Resources
-RB-level locations
.. .. .. No CSI-RS
REs
Contains
CSI-RS REs
1 slot time
0 1 2 3 4 5 6 7 8 9 10 11 12 13
CSI-RS 0
3000-3001 CSI-RS 1
CDM groups
3000-3001
PDCCH PDSCH
The DMRS is a UE-specific reference signal, which means that it is inserted before the
precoder and in principle only transmitted when and where the intended UE has DL-
SCH resources allocated.
The UE-specific reference signals eliminates the overhead of transmitting additional
cell-specific reference signals on the added antenna ports to support a larger number
of ports.
In this figure we see an example of how DMRS and CSI-RS might be distributed in
one slot (14 OFDM symbols).
1 slot time
0 1 2 3 4 5 6 7 8 9 10 11 12 13
CSI-RS 0
subcarriers
3012-3015 CSI-RS 1
CDM groups
3008-3011 DMRS for PDSCH
3004-3007
3000-3003 3000-3003
PDCCH PDSCH
The DMRS is a UE-specific reference signal, which means that it is inserted before the
precoder and in principle only transmitted when and where the intended UE has DL-
SCH resources allocated.
The UE-specific reference signals eliminates the overhead of transmitting additional
cell-specific reference signals on the added antenna ports to support a larger number
of ports.
In this figure we see an example of how DMRS and CSI-RS might be distributed in
one slot (14 OFDM symbols).
— Comb 2 or 4
subcarriers SRS
4 Combs
Intra- and inter-slot hopping with N = 2 and r = 1
symbols
Example: 4-symbol SRS with comb-4
© Ericsson AB 2018 | | LZU1082750 PA1 | Figure 3-64
Sequence generation
NR SRS reuses the LTE non-resource specific Zadoff-Chu based sequences
Sequence initialization formula includes both slot and symbol indices
Design parameters
Comb levels are 2 and 4
Number of SRS ports are 1, 2 or 4 per SRS resource
Cyclic shift separation between ports on the same comb
Maximum of 12 cyclic shifts for comb 4
Maximum at least 8 cyclic shifts for comb 2
Group and sequence hopping
30 sequence groups with 1 or 2 root-sequences per group
Group and sequence hopping are mutually exclusive
If sequence hopping is enabled, group hopping is disabled
Sequence ID is configured in UE specific manner using RRC
Bandwidth
Minimum BW is 4 PRBs and maximum BW is 272 PRBs
64 BW configurations (multiples of 4PRBs) in a single table
For BW <=96 PRBs: 17 unique configurations from LTE tables and 8
new configurations;
For 96< BW <=272: 39 new configurations
Frequency hopping and repetition
Intra-slot and inter-slot frequency hopping within a BWP
May be combined with repetition within the slot, with factor
r = 1, 2, 4 and r ≤ N (number of OFDM symbols in resource)
Same hopping formula as in LTE, but with a symbol-based counter function of
N and r
Frequency domain starting position
Frequency domain starting position is UE-specifically configurable to align on
a common grid with 4 PRB granularity
SRS
Comb 2 over one symbol Comb 4 over four symbols
— SRS resource set, each SRS set contains one or more SRS:s
Mapping to antennas
Device F1
SRS #1
SRS #2
F2
— RX beam sweep
— Single SRS port transmitted over
consecutive OFDM symbols
— Frequency hopping
— Frequency sweep of same SRS port over
consecutive OFDM symbols
In this table we see the antenna ports introduced in Rel 8, 9 and 10. There
are 8 ports (7-14) for DMRS and 8 ports for CSI-RS (15-22). More ports are
added in Rel 13-14.
Two antenna ports are said to be quasi co-located (QCL) if the large-scale
properties of the channel over which a symbol on one antenna port is
conveyed can be inferred from the channel over which a symbol on the
other antenna port is conveyed. The large-scale properties include one or
more of delay spread, Doppler spread, Doppler shift, average gain, and
average delay.
Summary Chapter 4
4 Detail the physical procedures for user data transmissions
4.1 Detail the reference signals related to user plane transmissions (DMRS, CSI-RS, TRS, PTRS, SRS)
4.2 Explain Type A and Type B transmissions
4.3 Explain HARQ codebook principles and Code Block Group (CBG) based retransmissions
4.4 Describe UL and DL scheduling principles and resource allocation in frequency and time domains
4.5 Explain link adaptation principles
4.6 Describe LDPC channel coding
′ ′ ′ 24
′
CB CB
Code block Code block Code block CRC
CRC
CB CB
Code block Filler Bits Code block Filler Bits Code block CRC
Filler Bits
CRC
22 for BG1 0
1
{2, 4, 8, 16, 32, 64, 128, 256}
{3, 6, 12, 24, 48, 96, 192, 384} 22 for BG1
2 {5, 10, 20, 40, 80, 160, 320}
10, 9,8,6 for BG2 3
4
{7, 14, 28, 56, 112, 224}
{9, 18, 36, 72, 144, 288}
10 for BG2
5 {11, 22, 44, 88, 176, 352}
6 {13, 26, 52, 104, 208}
© Ericsson AB 2018 | | LZU1082750 PA1 | Figure 3-72
7 {15, 30, 60, 120, 240}
LDPC – CB segmentation
0.95
BG1 BG1 BG1
1 1 8448
′
/ 24 1
0.67
′ 24 /
BG2
0.25 1
′ BG2 / 24 1
3840 ′ 24 /
16 24
308 3840 3848 8448
From base
column graph to parity check matrix • 2 BGs
• BG1: 46x68
0 0 0
R=2/3 R=1/2 R=1/3 R=1/4 R=1/5
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52
1 1 1 1 1 1 1 1 1
2 1 1 1 1 1 1 1 1 1 1
• BG2: 42x52
3 1 1 1 1 1 1 1 1
0
4 1 1 1 1 1 1 1 1 1 1
5 1 1 1 1
0 0 0
6 1 1 1 1 1 1
7 1 1 1 1 1 1
• 51 values of per BG
8 1 1 1 1 1 1
9 1 1 1 1
10 1 1 1 1 1
11 1 1 1 1 1
row
12 1 1 1 1 1
• Largest 384
13 1 1 1 1
14 1 1 1 1 1
15 1 1 1 1 1
0
16 1 1 1 1
•
18 1 1 1 1 1
19 1 1 1 1
0 0 0 0
20 1 1 1 1
21 1 1 1 1
0 0 0
25 1 1 1 1
0
26 1 1 1
17664 x 26112
27 1 1 1 1 1
28 1 1 1
29 1 1 1 1
30 1 1 1
31 1 1 1 1 1
1 0 0 0
Z ,
0 1 0 0
For , 0 I( Pi , j ) I
0 0 1 0
0 0 0 1
mod ,
, , Identity matrix of size
0 1 0 0
0
0 0 0
For , 0 I( Pi , j )
,
0 0 0 1
1 0 0 0
- Shortening (i.e. assigning info bits to known values, e.g. 0) is used to provide info
block size flexibility (R1-1701254)
Rate matching Note: The illustration here is for the case when LBRM is not
applied. When LBRM is applied, the circular buffer may shrink,
and the RV positions are scaled proportionally.
56·Z
3rd transmission
17·Z
RV0
1st transmission
RV 3
33·Z
2nd transmission RV1
Circular buffer of BG1
RV2 0
43·Z
4th transmission
13·Z
25·Z
© Ericsson AB 2018 | | LZU1082750 PA1 | Figure 3-76
Circular buffer of BG2
Rate Matching
, , ,
1 or 2
LDPC encoding
parameters
and PCM LDPC encoding
Scheduler Core
,
MCS Base graph 1 or 2
determination
TB size determination Code Block
Rate
Segmentation
matching
Resource allocation
TB CRC
Attachment
,
CBGTI
Rate matching output
© Ericsson AB 2018 | | LZU1082750 PA1 | Figure 3-78 Sequence length
determination
Scrambling
— Length-31 Gold sequence, same as LTE, is used for PDSCH data scrambling
Scrambling , ,…,
, ,…,
mod 2
2 2