GPRS Eng

GPRS
General Packet Radio Service
Helmut Visel ACTERNA Arbachtalstr. 6, D-72800 Eningen u.A.

Tel. +49 7121 - 86 2132 Fax +49 7121 - 86 2145 e-mail: [email protected] http://www.acterna.com
Contents
1. Introduction
-
3 Generations of Mobile Communication Evolution of Bandwidth The Mobile Internet Whats In It For Me? Features and Properties of GPRS The Old GSM World Cellular Network Structure GSM Interfaces Finding the Mobile Subscriber GMSK Modulation FDMA-TDMA Access Using GSM 900 as an Example TDMA Frame Hierarchy Burst Mask Logical GSM Channels GSM Channel Mapping Normal Burst Synchronization Burst Access Burst Frequency correction Burst Dummy Burst Checking the Subscriber Identity Data Ciphering Channel Coding Power Regulation Timing Advance Maximum Cell Radius? Frequency Hopping Interleaving What Happens After a Mobile Station is Switched On? IMSI Attach
2.
The world of IP
-
The TCP/IP Protocol Family Progress from a Message to a Full Packet Connection-oriented and Connectionless Protected and Unprotected Transmission The MAC Address The IP Address MAC and IP Address Internet Protocol (IP) IPv4 Frame Format IPv6 Frame Format Bridge / Switch / Router The Task of the Router
ACTERNA
Page 1
GPRS-contents.doc
When Does a Router Route? The Transport Layer The Port, the Application Address TCP Transmission Process TCP/UDP Multiplex Mechanism TCP Connection TCP Frame Format UDP Frame Format Domain Name System
3.
From GSM to GPRS Features and Basic Processes

-
GSM Network Structure BSS Aspects GPRS Network Extensions, SGSN, GGSN GPRS Network Structure Bandwidths Important GPRS System Requirements Fundamental Concepts Reference Points and Interworking High Level Functions Logical Architecture, Tasks of SGSN and GGSN GPRS Backbone Networks Mobility Management Interactions Between SGSN and MSC/VLR GPRS Attach Procedure GPRS Detach Procedure Security Functions Location Management Classmark Handling Radio Resource Functions Packet Routing and Transfer: Functions Packet Routing and Transfer: PDP States Packet Routing and Transfer: PDP Context Activation Packet Routing and Transfer: PDP Context Modification Packet Routing and Transfer: PDP Context Deactivation MM and PDP Context Transmission: Layer model Transmission: LLC Functions Transmission: Subnet Dependent Convergence Packaging the Protocols at the Air Interface Identifiers Quality of Service Profile Quality of Service Negotiation GPRS Terminal Equipment Classes Charging
4.
GPRS Radio Interface
ACTERNA
Page 2
GPRS-contents.doc
Um Interface Data TDMA Frame Hierarchy Physical Channels Logical Channels Multiframe Structure of a PDCH Downlink Mapping Uplink Mapping Timeslot Mapping for Compact Cells Compact Downlink Mapping Compact Uplink Mapping Principle of Radio Resource Management Resource Distribution Layer Model at the Um Interface Radio Block Structure Channel Coding Cell Reselection Timing Advance Power Control Activities During TA and IDLE Frames
5.
Radio Link Control / Medium Access Control

-
Layer Model Protocol Architecture RLC/MAC: Frame Structure for User Data RLC Modes RLC/MAC: Frame Structure for Signaling Multiplex Principle Packet Idle Mode and Packet Transfer Mode Mobile Originated Packet Transfer Packet Uplink Assignment Example of a Fixed Allocation Uplink RLC Data Block Transfer Mobile Terminated Packet Transfer Downlink RLC Data Block Transfer Packet Downlink Assignment Frequency Parameters Broadcast System Information Measurement Reports Cell Reselection Paging Procedure PDCH Release Procedure RLC/MAC Messages
6.
Protocols Between BSS and SGSN

-
Layer Model Network Service: Internal Architecture Network Service: Addressing
ACTERNA
Page 3
GPRS-contents.doc
Network Service: Load Sharing Network Service: VC Management Function Network Service: Service Primitives Network Service: Sub-Network Network Service: Frame Relay Network Service: Frame Relay Interface Network Service: Frame Relay Frame Structure Network Service: Frame Relay Routing Network Service: Frame Relay Parameters Network Service: Frame Relay Signaling Network Service: Control Protocol Network Service: Encapsulation Network Service: Control PDUs BSSGP Service Model The BVCI, LSP and NSEI Routing Parameters Services on RL and BSSGP SAPs DL/UL Transfer on RL and BSSGP SAPs Services on SAP GMM Paging Procedure on SAP GMM Services on NM SAP Frame Structure of the Protocol Messages PDU Definitions
7.
Communication Between MS and SGSN

-
Layer Model LLC Layer Reference Model LLC Features LLC Acknowledged and Unacknowledged Modes Data Link Connection Identifier (DLCI) LLC Layer, Structure and SAPs LLC Frame Structure Information Transfer Format (I-Format) Supervisory Format (S-Format) Selective Acknowledgement Frame (SACK Frame) Unconfirmed Information Format (UI-Format) Unnumbered Format (U-Format) Exchange Identification Frame (XID) Negotiable LLC Parameters Overview of Frame Types TLLI Assignment ABM Establishment Procedure Negotiating LLC Parameters LLC Parameters GMM Reference Model SNDCP Protocol Stack Multiplexing Different Protocols PDP Multiplexing The SAPs Around SNDCP
ACTERNA
Page 4
GPRS-contents.doc
SNDCP Model Segmenting Compression Negotiating Layer 3 Parameters Acknowledged Data Transfer Unacknowledged Data Transfer SN-PDU Format TOM Envelope Format GPRS SMS Transfer
8.
The Tunnel Principle and GPRS Tunneling Protocol

-
Protocol Stack The Principle of the Tunnel GPRS Tunnel Principle GPRS Tunneling Protocol (GTP) GTP Header GTP Messages Overview of GTP Messages Routing and Context Parameters Context Activation Context Deactivation
9.
Glossary
10. Data Sheets and Application Notes
ACTERNA
Page 5
GPRS-contents.doc
Chapter 1 Chapter 1
Introduction
Those who know nothing have to believe everything

(Marie von Ebner-Eschenbach)
Page 1-1
Notes:
ACTERNA
Introduction
Page 1-1
3 Generations of Mobile Communication 3 Generations of Mobile Communication
Erste Generation
Zweite Generation
GPRS GPRS
Dritte Generation
Analoge Analoge Mobiltelefonie Mobiltelefonie
Digitale Digitale Mobiltelefonie Mobiltelefonie
UMTS UMTS
1957
1972
1985
1992
2002
Page 1-2
1. Generation Analoge Mobilfunknetze. Diese sind in vielen Teilen der Welt weiterhin blich, bieten jedoch nicht die Mglichkeiten digitaler Netze. Der Datenverkehr mu per Modem dem analogen Kanal angepasst werden, dies ist umstndlich und verlustbehaftet. Insgesamt nimmt der Datenverkehr berall viel strker zu als der Sprachverkehr. In Deutschland waren das A-, B- und C-Netz Vertreter eines analogen Mobilfunknetzes. 2. Generation Digitales Netz. Der Teilnehmer bekommt eine einheitlich gute Sprachqualitt und zahlreiche Datendienste wie Fax, SMS und PC-Anbindung angeboten. Der Vorteil besteht weiterhin aus digitaler Signalisierung, die den Zustand und die Wnsche des Teilnehmers vielfltig und einheitlich wiederspiegelt. Rufumleitung und internationales Roaming sind Merkmale digitaler Netze. Das bisherige GSM-Netz ist ein Vertreter der zweiten Generation. 3. Generation Hier findet eine Erweiterung des digitalen Mobilfunknetzes zu breitbandigen Anwendungen hin statt. Bei UMTS ist dies zuknftig der Fall. Als bergang von der zweiten zur dritten Generation kann GPRS angesehen werden. Hier sind im Gegensatz zum alten GSM-Netz bereits weitaus hhere Bandbreiten mglich.
ACTERNA
Introduction
Page 1-2
Evolution of Bandwidth Evolution of Bandwidth

bertragungsgeschwindigkeit EDGE 384 kbit/s GPRS 62,4 kbit/s HSCSD Kanalbndelung 57,6 kbit/s Daten ber Sprachkanle 9,6 kbit/s Sprache UMTS max. 1,92 Mbit/s
1995
1996
1998
2000
2001
2003
Zeit
Page 1-3
EDGE (Enhanced Data Rates for Global Evolution) ist ein Datenbeschleuniger fr GSM und TDMA-Netze. Hier wird die Kapazitt der Zeitschlitze auf 48 kbit/s erhht, indem statt GMSK-Modulation eine 8-PSK-Modulation eingesetzt wird. Je Frequenz lassen sich dann 8 x 48 kbit/s, also 384 kbit/s bertragen.
ACTERNA
Introduction
Page 1-3
The Mobile Internet The Mobile Internet

Two worlds joining together ...
http://acterna.com
Application Application
Lower protocol layers
Lower
Mobile radio network

Gateway
protocol
Internet
layers
Page 1-4
The Internet on one side and the current GSM mobile radio network on the other side. One packet-oriented, the other circuit-oriented. The aim now is to join these two worlds together. The first thing is to extend the circuit-oriented GSM mobile radio network into a packet data network. This is done by introducing a few new network components and a set of special protocols into the GSM network. This gives users the opportunity to use the Internet, send e-mails and so on in the same way as with the fixed network, but while on the move. As far as the applications are concerned, theres no difference between the fixed and mobile networks. You could say that the application doesnt even see the mobile network.
ACTERNA
Introduction
Page 1-4
Whats In It For Me? Whats In It For Me?
GPRS offers mobile packet-oriented access to Packet Data Networks High bandwidth (up to about 170 kbit/s) and moderate, volume-dependent
tariffs offer advantages to the user.
Internet
such as the Internet, Intranet or X.25 Networks.
GPRS network X.25 network
e.g. Bluetooth interface
Page 1-5
The rising demand for data communications in both fixed and mobile networks has demonstrated the limitations of the Global System for Mobile Communication (GSM). It is capable of transmitting a maximum of 14.4 kbit/s within a TDMA timeslot. GSM Phase 2+ defines High Speed Circuit Switched Data (HSCSD). In HSCSD, up to eight TDMA timeslots are combined logically. Channel assignment is dynamic. This allows transmission rates to be increased to 115.2 kbit/s. Both techniques - GSM and HSCSD are less suitable for bursty traffic, such as WWW browsing. Customers are also not prepared to pay for such applications on a per time basis, desiring charges to be based on traffic volume instead. For these reasons, the European Telecommunications Standards Institute ETSI has standardized a new data communications service within GSM Phase 2+: the General Packet Radio Service GPRS.
ACTERNA
Introduction
Page 1-5
General Packet Radio Service (GPRS) This is an international standard. In theory, the radio data technology in existing GSM mobile radio networks can provide users with data rates of 171 kbit/s. In practice, the upper limit is currently around 50 kbit/s. GPRS is a packet-switched service for mobile data communications. The technique splits the information to be transmitted into small packets and transmits these independently of one another. The packets are reassembled in the correct order only in the receiver. This method optimizes the use of the available bandwidth of the mobile radio cell. Being packet-oriented, GPRS is ideal for IP applications. Mobile access to the Internet is therefore one of the important user benefits of the service. A further innovation for customers is the facility for billing according to the volume of data transmitted rather than only according to the duration of the call. The new GPRS technology is expected to result in increased utilization of mobile Internet access, mobile e-commerce and other mobile services such as the Short Message Service (SMS) and Wireless Application Protocol (WAP).
ACTERNA
Introduction
Page 1-6
Features and Properties of GPRS Features and Properties of GPRS
Data is transmitted as connectionless packets Up to about 170 kbit/s per user possible Simultaneous transmission of voice (GSM) and data (GPRS) User pays for bandwidth used, not for duration of call Permanent connection to the GSM network is possible Eliminates call set up for data transmission
Page 1-7
GPRS uses the same frequencies as GSM uses for voice transmission. The 200 kHz frequency band is subdivided into eight timeslots for transmitting voice signals. GPRS can combine these timeslots. This means that every mobile can use as much bandwidth as is needed for the current transmission. A protocol ensures fair access to the entire bandwidth. GPRS can be visualized as a large layer 2 access network to packet-switched data networks such as X.25 and IP networks. In this sense, GPRS is only one possibility of accessing data applications from a mobile terminal. Such networks are both cost-effective and user-friendly, thanks to volume-based charging, low waiting times and high speeds.
ACTERNA
Introduction
Page 1-7
GPRS terminals GPRS terminals
Page 1-8
Two different types of GPRS mobile stations.
ACTERNA
Introduction
Page 1-8
The Old GSM World The Old GSM World

A Review
Page 1-9
Notes:
ACTERNA
Introduction
Page 1-9
The Old GSM World The Old GSM World
X.25
OMC OMC
X.25 MSC MSC
MS
BTS
BSC
Abis
Base Station Subsystem (BSS)
MSC MSC
GMSC GMSC HLR VLR EIR AUC Network Switching Subsystem (NSS)
ISDN
Um 900/1800/1900 MHz FDMA / TDMA 8 Timeslots, 271 kbit/s (33.9 kbit/s per timeslot)
Page 1-10
A GSM system consists of individual subsystems that together make up a PLMN (Public Land Mobile Network). A distinction is made between the terminal, i.e. the mobile phone, and the base station subsystem (BSS) and network switching subsystem (NSS). The BSS basically covers the radio side of things. The NSS takes care of the switching. Together, the BSS and NSS have the task of guaranteeing the mobility of the terminal within the network or the different GSM networks of various network operators. The GMSC forms the gateway to the public circuit-switched fixed network (e.g. ISDN). MS BTS BSC MSC GMSC HLR VLR EIR OMC ISDN = = = = = = = = = = Mobile Station Base Transceiver Station Base Station Controller Mobile Switching Center Gateway Mobile Switching Center Home Location Register Visitor Location Register Equipment Identity Register Operation and Maintenance Center Integrated Services Digital Network
ACTERNA
Introduction
Page 1-10
Cellular Network Structure Cellular Network Structure

F2 F7 F1 F6 F5 F2 F7 F1 F6 F5 R D=5R F4 F3 F6 F5 F7 F1 F4 F4 F2 F3 F3
Page 1-11
Simulations and experiments have shown that a signal to noise ratio (S/N) of about 18 dB is sufficient for acceptable intelligibility of speech signals. S/N = Signal / (Noise + Neighbor cell interference) Assuming a circular propagation characteristic, this leads to a theoretical frequency repetition spacing D of about 4.6 R. Sectoring and the use of directional antennas currently means that the practical frequency repetition spacing can be as low as 1 R, i.e. the same frequency can be used again in the immediately adjacent cell.
ACTERNA
Introduction
Page 1-11
GSM Interfaces GSM Interfaces

Um G D
HLR
BTS
Abis B
VLR
C C
VLR
B E
BSC
A Ater
Signaling Signaling and data
MSC
GMSC
ISDN
EIR
Page 1-12
Apart from the Abis and Um interfaces, the GSM subsystems are connected together directly or indirectly using the worldwide SS #7 network. Consequently, the basis for all signaling (apart from on the Abis and Um interfaces) is the SS #7 signaling protocol. The connection to ISDN is formed by the ISDN User Part (ISUP). The Mobile Application Part (MAP) is available as the SS #7 user for data communications with HLR, VLR and EIR. Similarly, the GSM protocols via the A interface are transmitted using the MTP (Message Transfer Part) transfer mechanism of the SS #7 protocol. The services of the SCCP (Signaling Connection Control Part) are used by various subsystems in the GSM network. The Abis and Um signaling is based on LAP-D protocols similar to the familiar ISDN user signaling (D channel protocol).
ACTERNA
Introduction
Page 1-12
Finding the Mobile Subscriber Finding the Mobile Subscriber

MSRN
ISDN
1
VLR TMSI LAI

6
Paging
TMSI
8
MSISDN MSISDN
BTS
GMSC MSISDN
2 4
MSC
7
BSC
Location area
Paging
MSRN
3
MSRN TMSI
8
HLR
MSC
TMSI
BTS TMSI
Page 1-13
The dialed number (MSISDN) does not contain any information about the current location of the mobile subscriber. The location of the subscriber has to be determined first before the call can be routed. The MSRN is used for this. The MSRN is temporarily assigned to an MS by the currently appropriate VLR and entered in the HLR. The figure above shows the basic sequence of the routing process in the case of a mobile terminating call. The MS is only called in the location area where it is actually located.
MSISDN MSRN TMSI LAI
= Mobile Station ISDN Number = Mobile Station Roaming Number = Temporary Mobile Subscriber Identity = Location Area Identifier
ACTERNA
Introduction
Page 1-13
GMSK Modulation GMSK Modulation
f f + df
t f - df
MSK modulator MSK modulator

Gaussian filter B x T = 0.3
Page 1-14
The Gaussian minimum shift keying (GMSK) used for GSM/GPRS is a special variant of MSK (minimum shift keying). Frequency shift keying takes place relatively slowly at the output of the MSK modulator as a result of the Gaussian low-pass filter in the path of the modulation signal (data to be transmitted). This leads to a narrow frequency spectrum and hence to efficient usage of the radio resources. (The bandwidth of a frequency channel is 200 kHz). B = 3 dB bandwidth T= bit width (3.693 s)
ACTERNA
Introduction
Page 1-14
FDMA-TDMA Access Using GSM 900 as an Example FDMA-TDMA Access Using GSM 900 as an Example
f / MHz
Downlink
960
576.9 s
25 MHz
f 124 ... f2 f1 935
TS 1
TS 2
TS 3
TS 4 ...
TS 5
TS 6
TS 7
TS 8
45 MHz
TS 1 TS 1
TS 2 TS 2
TS 3 TS 3
TS 4 TS 4
TS 5 TS 5
TS 6 TS 6
TS 7 TS 7
TS 8 TS 8
4.615 ms
Uplink
200 kHz
915 TS 1 TS 2 TS 3 TS 4 ... TS 1 TS 1 890 t TS 2 TS 2 TS 3 TS 3 TS 4 TS 4 TS 5 TS 5 TS 6 TS 6 TS 7 TS 7 TS 8 TS 8 TS 5 TS 6 TS 7 TS 8
f 124 ... f2 f1
Page 1-15
A combination of FDMA and TDMA is used on the Physical Layer (Layer 1 of the OSI model) for multiple access in GSM. Two frequency bands separated by 45 MHz are reserved for GSM operations: the 890 MHz - 915 MHz range as uplink and the 935 MHz - 960 MHz range as downlink. Each of these 25 MHz wide bands is divided into 124 individual channels separated by 200 kHz. A guard band of 200 kHz remains in both the uplink and the downlink bands. Each Radio Frequency Channel (RFCH) has a unique number. Matching pairs of numbers from the uplink and downlink form a duplex channel with 45 MHz duplex spacing. Each of these 200 kHz channels contains 8 TDMA channels through subdivision into 8 timeslots each. The 8 timeslots of these TDMA channels are collected together into a TDMA frame. The TDMA frames of the uplink are transmitted with a delay of 3 timeslots relative to the downlink. A mobile station uses the same channel in the uplink and the downlink, i.e. the timeslot with the same number. It thus does not need to transmit and receive simultaneously and does not require a duplex unit, thanks to the three timeslot offset. This reduces the complexity of the mobile, resulting in lower production costs.
ACTERNA
Introduction
Page 1-15
TDMA Frame Hierarchy TDMA Frame Hierarchy

Hyperframe
0 1 2 3 4 2045 2046 2047
0 0
1 1
3 2
48 24
49
50
Superframe
25
24 25
49
50
Multiframe
2
Burst
TDMA frame
Page 1-16
TDMA frame: 8 timeslots Period = 4.615 ms Timeslot duration = 0.577 ms Multiframe: 26 frame multiframe for user data 51 frame multiframe for signaling Superframe: 51 x 26 frame multiframe or 26 x 51 frame multiframe Period = 6.120 s Hyperframe: 2048 superframes or 2715610 TDMA frames Period = 3 h 28 min 53.760 s Period = 120 ms Period = 235.38 ms
ACTERNA
Introduction
Page 1-16
Burst Mask Burst Mask
dB +4 -6
Tolerance mask 542.8 s (148 bits)
- 30
- 70
577 s (156.25 bits)
Page 1-17
Because the mobile always occupies the same assigned timeslot, it operates in burst mode.It sends a data packet in the direction of the BTS every 4.615 ms. The MS must pulse its output power to avoid interfering with adjacent timeslots. A level range of about 70 dB is required between the on and off states of the output stage. There are 114 bits available for transmitting ciphered speech/data within a normal burst. The remaining bits are used for synchronization and signaling.
ACTERNA
Introduction
Page 1-17
Logical Channels Logical Channels

Log. Channels Traffic Channel (TCH) Control Channel (CCH)
Full rate
Half rate
Broadcast (BCCH)
Common (CCCH)
Dedicated (DCCH)
Speech 22.8 kbit/s
Speech 11.4 kbit/s
Frequency Correction (FCCH)
Random Access (RACH)
Stand Alone (SDCCH)
Data max. 9.6 kbit/s
Data max. 4.8 kbit/s
Broadcast Control (BCCH)
Access Grant (AGCH)
Fast Associated (FACCH)
Synchronisation (SCH)
Paging (PCH)
Slow Associated (SACCH)

= Uplink = Downlink
Page 1-18
The radio transmission resources are divided into various logical channels. The first group is used for transmitting the user data (traffic channels). The second group, the control channels, is used to transmit widely differing types of signaling information. Broadcast Control Channels transmit information that the MS uses to synchronize itself in time and frequency to the GSM network. The MS also receives all the important data about its own cell and the neighbor cells through the BCCH. Common Control Channels are used by the MS to access the network. The network itself sends paging messages and assigns a Stand Alone Dedicated Control Channel to the MS for handling further signaling (in the event of a mobile originated call) over these channels. The Dedicated Control Channels are assigned to a particular mobile station by the network for handling the signaling. Once the MS has received a traffic channel, a SACCH or FACCH in association with this traffic channel is also available simultaneously. SMS data are transmitted over a SDCCH.
ACTERNA
Introduction
Page 1-18
Channel Mapping (Downlink) Channel Mapping (Downlink)

FN 0 1 2-5 6-9 10 11 12 - 19 20 21 22 - 25 26 - 29 30 31 32 - 35 36 - 39 40 41 42 - 45 46 - 49 50 Timeslot 0 FCCH SCH BCCH AGCH / PCH FCCH SCH AGCH / PCH FCCH SCH SDCCH/4 [0] SDCCH/4 [1] FCCH SCH SDCCH/4 [2] SDCCH/4 [3] FCCH SCH SACCH [0] 44 - 47 SACCH [1] Idle 48 - 50 Idle 25 Idle SACCH [3] 28 - 31 32 - 35 36 - 39 40 - 43 SDCCH/8 [7] SACCH [0] SACCH [1] SACCH [2] FN 0-3 4-7 8 - 11 12 - 15 16 - 19 20 - 23 24 - 27 Timeslot 1 SDCCH/8 [0] SDCCH/8 [1] SDCCH/8 [2] SDCCH/8 [3] FN SDCCH/8 [4] SDCCH/8 [5] Timeslot 2 - 7
Example of a logical channel configuration on the air interface
51 frame multiframe
26 frame multiframe
SDCCH/8 [6]
0 - 11
TCH
12
SACCH
13 - 24
TCH
Page 1-19
This diagram represents a possible channel configuration in the downlink. The BCCH is in timeslot 0. As the BCCH contains further information about the remaining channel configuration, it must always be located at a fixed position in the mapping. In this example, timeslot 1 is also occupied by signaling channels. Timeslots 2 through 7 are available for traffic channels and slow or fast associated control channels.
ACTERNA
Introduction
Page 1-19
Channel Mapping (Uplink) Channel Mapping (Uplink)

FN 0-3 4-5 6-9 Timeslot 0 SDCCH/4 [3] RACH SACCH [2] SACCH [3] 15 - 18 19 - 22 14 - 36 RACH 23 - 26 27 - 30 31 - 34 37 - 40 41 - 44 45 - 46 47 - 50 SDCCH/4 [0] 35 - 38 SDCCH/4 [1] RACH Idle 39 - 42 43 - 46 47 - 50 SDCCH/8 [5] SDCCH/8 [6] SDCCH/8 [7] SACCH [0] 25 Idle SDCCH/8 [0] SDCCH/8 [1] SDCCH/8 [2] SDCCH/8 [3] SDCCH/8 [4] FN 0-3 4-7 8 - 11 12 - 14 10 - 13 Timeslot 1 SACCH [1] SACCH [2] SACCH [3] Idle FN Timeslot 2 - 7
Example of a logical channel configuration on the air interface
51 frame multiframe
26 frame multiframe
0 - 11
TCH
12
SACCH
13 - 24
TCH
Page 1-20
There are, of course, no BCCH, no synchronization channels and no paging channel in the uplink mapping. Instead, timeslot 0 is largely occupied by the Random Access Channel, which the MS uses to signal itself to the network.
ACTERNA
Introduction
Page 1-20
Normal Burst Normal Burst
577 s
3 tail bits
57 bits Encrypted Information
1 S bit
26 bits Training Sequence
1 S bit
57 bits Encrypted Information
3 tail bits
Guard Time 8.25 bits (30.5 s)
Page 1-21
A normal burst is used to transmit practically all data between the mobile and the BTS. The only exceptions are the channel request of the MS in the Random Access Channel (RACH) and the transmission of the BTS synchronizing information in the synchronization channel (SCH). All other information in the control and traffic channels is transmitted as normal bursts in the uplink and downlink. The 114 bit channel coded user data already contains redundant bits as a result of channel coding and must not therefore be confused with the net payload bit rate. The stealing bit (Sbit) indicates to the receiver whether the payload information is traffic or FACCH data. 3 tailing bits at the start and end of the burst provide a time tolerance range that ensures that no data is lost if the burst does not exactly meet the timeslot for some reason. The receiver looks for the known training sequence within the burst. If this is found, it can determine where the user data is located and compensate for interference and multipath propagation delays of up to 16 s. The transmitter of a burst must switch over the power level of its output stage. The necessary settling takes place during the guard time.
ACTERNA
Introduction
Page 1-21
Synchronization Burst Synchronization Burst
577 s
3 tail bits
39 bits SCH data
64 bits Extended Training Sequence
39 bits SCH data
3 tail bits
Guard Time 8.25 bits (30.5 s)
Page 1-22
The synchronization burst is used to transmit information in the synchronization channel (SCH). The receiver can already determine whether the transmitter is a GSM system or not by means of the over-length training sequence. The 78-bit SCH data contains synchronization information that allows the MS to time synchronize to the mobile radio network. Among other things, the SCH bits contain the absolute frame number (FN) as well as the base station identity code (BSIC) of a BTS for initial rough identification of a cell.
ACTERNA
Introduction
Page 1-22
Access Burst Access Burst
577 s
8 tail bits
41 bits Synchronization Sequence
36 bits Data (RACH)
3 tail bits
Guard Time 68.25 bits
Page 1-23
The access burst is used by the mobile for the initial access to the BTS in the Random Access Channel (RACH). As the MS has no information about the timing advance (TA) at this point, the MS transmits the burst without a time lead. To ensure that the burst at the receiving end remains within the timeslot, it is particularly short and has an extremely long guard time. The BTS can determine the timing advance from the time offset at the receiving end and then transmit this information back to the mobile. The mobile uses the access burst for a mobile originating call and to request a channel during a handover.
ACTERNA
Introduction
Page 1-23
Frequency Correction Burst Frequency Correction Burst
577 s
148 bits Zero Sequence
Guard Time 8.25 bits
Page 1-24
The frequency correction burst is only broadcast in the Frequency Correction Channel (FCCH) in the downlink. All 148 bits are set to 0. In conjunction with the GMSK modulator, this zero sequence gives a constant line of 67.7 kHz above the BCCH center frequency in the frequency spectrum. This line is used by the mobile to synchronize to the BCCH frequency of a BTS.
ACTERNA
Introduction
Page 1-24
Dummy Burst Dummy Burst
577 s
3 tail bits
142 bits Fill data (pseudo-random bit sequence specified by GSM)
3 Guard Time tail 8.25 bits bits
Page 1-25
After switch-on, the mobile uses the BCCH frequencies of the adjacent BTSs as reference in the selection of the serving cell. The level measurements on which handover decisions are based are also made on the BCCH frequencies. It is therefore necessary that the BCCH TRX always transmits at a constant level in all timeslots (no level regulation) and that no empty timeslots are present. Empty timeslots in the BCCH frequency are therefore filled with the dummy burst in contrast with the other carrier frequencies.
ACTERNA
Introduction
Page 1-25
Checking the Subscriber Identity Checking the Subscriber Identity

Authentication
BTS
AUTH_REQ RAND
Ki
A3
SRES
RAND
AUTH_RSP SRES
Page 1-26
Ki RAND SRES A3
Individual value unique to each SIM card Random number generated by AuC (Signed Response) result of authentication test Algorithm used to calculate SRES
SRES = f(Ki, RAND)
ACTERNA
Introduction
Page 1-26
Data Ciphering (1) Data Ciphering (1)

Ciphering
BTS
CIPH_MODE_COMMAND A5/x (x = 1...5)
Ki
A8
Kc
RAND
CIPH_MODE_COMPLETE
The MS calculates its ciphering key from Ki and RAND
Page 1-27
Procedure for (activating) data ciphering on the air interface after the preceding authentication test. Kc = Ciphering key A8 = Algorithm used to calculate Kc Kc = f(Ki, RAND) The CIPH_MODE_COMPLETE message is already ciphered when transmitted.
ACTERNA
Introduction
Page 1-27
Data Ciphering (2) Data Ciphering (2)

Ciphering
FN = f(t) Kc
A5/x
Ciphering sequence (114 bits)
1100100010001.......11100001010011
114 bits non-ciphered data
XOR
0001001001111......0101000001011
114 bits ciphered data
Burst
Page 1-28
FN Kc A5/x
= = =
Frame number Ciphering key Algorithm used to calculate the ciphering sequence (5 different algorithms are available)
ACTERNA
Introduction
Page 1-28
Channel Coding for Signaling Data Channel Coding for Signaling Data
184 bits signaling data
Block coding (Fire code) 184 bits

40 parity bits 4 tail bits
Convolutional (hash) coding
456 bits coded signaling data
Page 1-29
The highly variable nature of a mobile radio channel results in very high bit error rates of some 10-3 to 10-1, which is unacceptable for meaningful transmission of speech or data signals (to say nothing of transmitting signaling information). This problem is tackled by appropriate channel coding. Sufficient redundancy is added to the bit stream prior to transmission to ensure that transmission errors can be detected and corrected automatically without the need to repeat frames. Using this method, error rates of between 10-5 and 10-6 can be achieved. Block coding is used first of all to add parity check bits and fill bits to the raw data. The code this produced is an extremely powerful Fire code. The data block this produces is then subjected to convolution or hash coding which adds further necessary redundancy.
ACTERNA
Introduction
Page 1-29
Channel Coding for Speech Data Channel Coding for Speech Data
260 bits from speech coder (260 bits / 20 ms = 13 kbit/s) 50 class 1a bits 132 class 1b bits 78 class 2 bits
Block coding
Block coding
50 bits
3
parity
132 bits
4
def. bits
Convolutional coding
378 bits
78 bits
Page 1-30
In contrast with the signaling data channel coding, speech data coding uses a somewhat simplified procedure. Since the speech signal itself contains a high degree of redundancy, excessive error correction facilities are unnecessary. For example, certain bits in the speech signal are transmitted without any form of error correction. Errors in these bits do not cause a discernible reduction in the quality of the transmitted speech. Class 1a bits: Class 1b bits: Class 2 bits: Important bits where errors cause the most impairment Less important class 1 bits These unimportant bits are transmitted without any protective measures.
All class 1 bits are subjected to convolutional coding that offers a high degree of correctability in the event of transmission errors.
ACTERNA
Introduction
Page 1-30
Power Regulation Power Regulation

P(TX)
Timeslot 2
BTS
t
Ti m es lo t5
P(TX) P(TX)
TS 0 TS 1 TS 2 TS 3 TS 4 TS 5 TS 6 TS 7
Downlink
Uplink
Page 1-31
GSM defines that power regulation is to be performed for the mobile station, but it is optional for the BTS. Depending on the receive quality, the BSC can request the BTS and the mobile station to adjust their output power levels. Power regulation serves on the one hand to minimize interference impairments and also ensures longer-term constancy in the battery power supply of the mobile station on the other hand. All the downlink channels of the BCCH TRX are excluded from power regulation. These must always be operated at maximum output power, so that the BCCH frequency can be used as the reference frequency for cell selection and reselection. The figure above shows a frequency without BCCH for which power regulation is performed.
ACTERNA
Introduction
Page 1-31
Timing Advance Timing Advance
Large timing advance
BTS (RX)
Timeslot n Timeslot n+1
BTS
a (TX)
TA(a)
b (TX)
TA(b)
t Small timing advance
Page 1-32
To ensure that all the bursts from the various mobile stations always fit precisely into the timeslots of the TDMA frames at the BTS, those mobile stations located further away must transmit their data bursts somewhat earlier than the mobile stations that are located closer to the BTS. The different propagation delays for the bursts from mobile stations located at different distances from the BTS are equalized in this way. The corresponding derivative time value is continuously indicated to the MS from the network as the timing advance parameter. The timing advance resolution is about 550 m (value range TA = 0..63). This corresponds to the distance traveled by light in the time represented by a bit width (3.6 s).
ACTERNA
Introduction
Page 1-32
Maximum Cell Radius? Maximum Cell Radius?

Burst (TB = 325 s) Guard time (TG = 252 s)
Access burst:
Location (RX/TX)
BTS (RX)
Burst TB = 325 s TD
TS = 2TD + TB rmax = V0 x TD
MS (TX)
Burst TB = 325 s
TS = 577 s TB = 325 s V0 = 300 000 km/s
MS (RX)
TS 7
TS 0
TS 1 rmax = V0 x (TS - TB) / 2 = 37.8 km
TD BTS (TX) TS 7 TS 0 TS = 577 s TS 1 t
rmax = 35 km by definition
Page 1-33
The figure above clearly shows why the maximum cell radius for GSM (GSM 900 in the example here) shows such a sharp limit.. This is not due to any special feature of radio signal propagation, but it is strictly limited by the applicable GSM specifications (and by the speed of light).
ACTERNA
Introduction
Page 1-33
Frequency Hopping Frequency Hopping
TDMA frame 1
TDMA frame 2
TDMA frame 3
f3 TS 0 TS 1 TS 2 TS 3 TS 4 TS 5 TS 6 TS 7 TS 0 TS 1 TS 2 TS 3 TS 4 TS 5 TS 6 TS 7 TS 0 TS 1 TS 2 TS 3 TS 4 TS 5 TS 6 TS 7
Page 1-34
To counteract the problems of multipath propagation and the resultant fading effects, GSM optionally uses frequency hopping on the air interface. Signal drop outs caused by multipath propagations (e.g. reflections from a hillside or high building) are dependent on phase and therefore also on frequency. If the frequency is changed after each TDMA timeslot, this reduces the statistical probability of signal drop outs. The BTS and the mobile must agree on a mutual hopping sequence before the start of a connection.
ACTERNA
Introduction
Page 1-34
Interleaving Interleaving
Data blocks after channel coding

1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8
Page 1-35
Interleaving is another method used in addition to frequency hopping for improving the transmission quality over the air interface. To ensure that a brief impairment on the air interface does not result in the loss of an entire data block, the blocks ready for transmission are not sent in chronological order, but are distributed across several bursts. The same number of bits will be lost in the event of an impairment, whether interleaving is used or not. However, the lost bits will only be parts of complete data packets, which may still be usable. The degree of interleaving varies for speech and signaling data. Signaling data is spread out across more bursts than speech data.
ACTERNA
Introduction
Page 1-35
What Happens After a Mobile is Switched On? What Happens After a Mobile is Switched On?
Switch on mobile
The MS searches its radio environment for BCCH frequencies using the following criteria: BCCH frequencies are broadcast at constant level and have a constant amplitude spectral line 67.7 kHz above their center frequency.
The MS synchronizes itself to the BTS frequency that can be received at the best level.
The MS now monitors the synchronization channel (SCH) to synchronize its timing to the network (hyperframe, frame number).
After frequency and time synchronization, the mobile can extract all the data pertaining to the network, its own cell, and the neighbor cells from the SYS-INFO message elements that are broadcast periodically on the BCCH.
Mobile ready to use
Page 1-36
Unbelievably, you can actually use a mobile to make phone calls ...
ACTERNA
Introduction
Page 1-36
IMSI Attach IMSI Attach

The mobile registers itself ...
LOCATION UPDATE REQUEST { Ground = IMSI attach LAI IMSI or TMSI }
LOCATION UPDATE ACCEPT { LAI IMSI or TMSI }
MS
BTS
Page 1-37
The BTS continually transmits the parameter ATT in BCCH SYS-INFO 3. This indicates whether the IMSI Attach/Detach function is activated. This determines how the mobile is to react. If IMSI Detach is activated, the MS logs off from the network before it is switched off and the appropriate entry is made in the VLR. The advantage here is that no paging messages need to be sent to a mobile that is switched off, and the caller receives faster notification that the called party is currently unavailable. The complement of IMSI Detach is IMSI Attach. Here, the MS logs on to the network when it is switched on. IMSI Attach can be compared to the periodic location update. IMSI Attach uses the same procedures as are used for a location update, except that the cause in the LOCATION UPDATE REQUEST message is set to IMSI Attach. The current location area and subscriber identity are also transmitted.
ACTERNA
Introduction
Page 1-37
Chapter 2 Chapter 2
The World of IP
Page 2-1
Notes:
ACTERNA
The World of IP
Page 2-1
The TCP/IP Protocol Family The TCP/IP Protocol Family

Application layer HTTP HTTP FTP FTP SNMP SNMP Name Name Server Server NFS NFS
Transport layer
TCP TCP (Transmission Control Protocol) (Transmission Control Protocol)
UDP UDP (User Datagram Protocol) (User Datagram Protocol)
Internet layer
IP IP (Internet Protocol) (Internet Protocol)
Networks
ARPANET ARPANET
ATM ATM
X.25 X.25
Ethernet Ethernet
Token Ring Token Ring
Page 2-2
The TCP/IP family of protocols has become standard in the higher layers above LLC in almost all systems. The core covers layers three and four, with IP representing layer 3 and TCP/UDP layer 4. Applications are addressed directly above the TCP/UDP layer. The TCP/IP stack model was developed independently of the OSI model, which explains the absence of layers 5 and 6. The Internet Protocol (IP) has the following tasks: Fragmentation and reassembly of data packets Routing using the IP address and the routing tables in the routers Error treatment and indication (optional) The Transmission Control Protocol assumes the task of setting up the end-toend connection. In this sense, it provides a connection-oriented transport service. TCP guarantees delivery of the data packets in the correct order. Protection is provided by numbering the data packets and repeating lost or damaged packets. The data packets are assigned to the corresponding applications in the application layer by means of so-called ports. Fixed ports are assigned for Internet applications (e.g. 21 for FTP). The User Datagram Protocol provides a connectionless transport service. UDP does not guarantee delivery of a datagram to the destination partner, neither does it make any provisions to prevent the duplication or incorrect ordering of datagrams.
ACTERNA
The World of IP
Page 2-2
Progress from a Message to a Full Packet Progress from a Message to a Full Packet
Of course, every function (layer) has to add something
Message + Application info + Presentation info + Communication info + Port and poss. Transport info + IP info + MAC info
Data HTTP
TCP IP MAC Header FCS Trailer
= Data frame
MAC IP TCP HTTP Data FCS
Page 2-3
From layer 2 onwards, the individual layers according to the OSI model are concerned with organizational and logical tasks. The processing of these tasks is defined in protocols. Each data frame transported by the network contains information (header / trailer) from each layer that is involved. There is no rule that every function level must be used in data communications between two users. In our example, layers 5 and 6 are not used. TCP: Transport Control Protocol
HTTP: Hyper Text Transfer Protocol
ACTERNA
The World of IP
Page 2-3
Connection-oriented and Connectionless Connection-oriented and Connectionless

Properly connected or anyhow
Connection-oriented:
Info Address C
Info Address B Info Address A
S S
S S S S
Connectionless: R R
Info Address B Info Address B Info Address B
R R R R
B
Info Address B
Page 2-4
Connection-oriented: The connection must be set up before transmission takes place. The set up can take place dynamically by means of signaling (dialing) or statically by means of configuration. In classical telecom networks, a line is switched and connected between the two end users (circuit-switched). Technologies such as X.25, Frame Relay or ATM operate on a frame or cell basis. The cell or frame header is labeled to correspond with a virtual channel. This label changes from one path segment to another. The actual destination address of the receiver is not evident. During the set up process, the virtual channel labels from one connection node to the next are established. Note: The Transmission Control Protocol (TCP) provides a connection-oriented end-toend transport service. Connectionless: Every packet contains the complete destination address (IP address). Each router decides for itself how the packet is to be forwarded. The transmitter passes the data frame with complete destination address on to the medium (just anyhow) and relies on the network to forward the data frame to the destination. Note: The User Datagram Protocol (UDP) provides a connectionless end-to-end transport service.
ACTERNA
The World of IP
Page 2-4
Protected and Unprotected Transmission Protected and Unprotected Transmission

Often, there is no point in protection.
Protected Bidirectional communication File transfer Remote access Unprotected Unidirectional communication Multicast Real time e-mail
Page 2-5
Protected: File transfer and remote access are two examples where 100% control of the transport process is required. They are bi-directional, since the receiver must return an acknowledgement to the transmitter that the data have been received correctly. This exchange, transmitting data and returning acknowledgements, takes place continuously throughout the data transfer. Unprotected: It is simply impossible for each receiver to acknowledge receipt in multicast applications (where a transmitter broadcasts data simultaneously to several receivers). Unprotected transport is therefore used in such cases. In realtime applications (such as voice over IP), repeating faulty or lost packets makes no sense. E-mails usually consist of small quantities of data, and there is no point in setting up a protected transport connection, since this would result in an excessive and unnecessary overhead.
ACTERNA
The World of IP
Page 2-5
The MAC Address The MAC Address

00 - 01 - 68 - 50 - 23 - 45
Manufacturer code Card code Consists of a total of 6 bytes Examples: Unicast: 00-01-68-50-23-45 Broadcast: FF-FF-FF-FF-FF-FF Multicast: 01-80-C2-00-00-00 The MAC address is only a terminal address The country and area codes are missing
Page 2-6
To correspond with the different layers, a distinction is made in the network between the hardware address (MAC address, Media Access Control) and the IP address. The MAC address is a simple, globally unique terminal address which is only suitable for addressing within local area networks. The network addressing, necessary for structuring large networks by assigning several users to a certain network segment or Internet subnet, is missing. The Unicast address: is the individual users address. Each address may only occur once worldwide. The first 3 bytes determine the card manufacturer. The last 3 bytes are assigned by the card manufacturer, who must ensure that each combination is assigned once and once only. These burnt-in hardware network addresses are not normally used in token ring networks. If the address starts with 40 (hex), it is an administratively assigned address. A Broadcast address: consists of all ones in binary. This corresponds to 6 x FF in hexadecimal code. A Multicast address: always has a one as the first bit. Example: 01-80-C2-00-00-00 is written in canonical address format. Canonical means lowest significant bit (LSB) first. The canonical form is used for Ethernet MAC addresses. Token Ring and FDDI use the MSB format (most significant bit first). 01-80-C2-00-00-00 (LSB) => 80-01-43-00-00-00 (MSB)
ACTERNA The World of IP Page 2-6
The IP Address The IP Address

Addressing in a large (worldwide) network
Network ID Host ID
Network ID
Subnet ID
Host ID
141 . 169 . 30 . 117
Internet
ACTERNA Intranet
ACTERNA Eningen
My PC
Page 2-7
An Internet address has a fixed length of 4 bytes (Note: New version 6 IP addresses are 16 bytes wide!). These are represented by decimal numbers separated by dots. It is divided into a network part and a host part. The Network Information Center (NIC) assigns the addresses for the network part for all institutions and companies wishing to be connected in any way to the international Internet. The assignment of subnet and station addresses is taken care of by the organizations themselves. Thus, an address that is unique in the worldwide Internet is generated which is clearly assigned to an organization (ACTERNA), a subnet (ACTERNA Eningen) and a user (My pc). When IP network part addresses are assigned, it is actually address ranges that are assigned. There are 3 classes of address ranges, Class A, B and C. The address class determines the maximum number of possible subnet and host addresses. Class A networks: 16,7 million individual IP addresses Class B networks: 65,536 individual IP addresses Class C networks: 256 individual IP addresses The (local) IP network of an organization is called an Intranet.
ACTERNA
The World of IP
Page 2-7
MAC and IP Address MAC and IP Address

MAC for the local and IP for the worldwide network
IP host Data Router IP host Data
IP1
IP for worldwide network IP
IP2
MAC1
MAC for local network
MAC2
MAC for local network
MAC3
Data
IP2
MAC2
Data
IP2
MAC3
Page 2-8
Locally, within a broadcast domain, a receiver must be addressed using its MAC address. The MAC layer (Media Access Control) regulates access to the LAN. The IP address is media-independent.
ACTERNA
The World of IP
Page 2-8
Internet Protocol (IP) Internet Protocol (IP)

Internets need an Internet protocol
Internet protocol Network 1 Router Network 2 Router
Network 3
Layer 2 protocol
Page 2-9
The formation of IP networks requires a new addressing layer and therefore a corresponding protocol that supports the transport of packets via the routers into the destination network. IP (Internet Protocol): IP is a layer 3 protocol. It is a protocol for connectionless transport. IP transports packets over the best path through the network in each case. It does not provide protected transmission. It fragments the packets if they are longer than the packet length permitted for the transit or destination network. The layer 2 addressing only applies to the local subnet.
ACTERNA
The World of IP
Page 2-9
IP Version 4 (IPv4) Frame Format IP Version 4 (IPv4) Frame Format

Bit 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Version Header length Total length Identification F F F Time to live IP header checksum IP source address (32-bit) IP destination address (32-bit) Options Fragment offset Protocol Type of service
Datagram (max. 65536 bytes)
Page 2-10
Version indicates the IP version (e.g. Version 4 or Version 6) Header length specifies the total length of the IP header in units of 32 bits. Type of service This field specifies services and their network-relevant parameters (priorities, router resource requirements, etc.) Total length indicates the length of the datagram in bytes. TL is 16 bits wide, so the datagram can have a maximum length of 0xFFFF bytes. Identification is used to uniquely identify individual data packets. F The 3 flag bits indicate whether the datagrams are fragmented and therefore require reassembly in the receiver. Fragment offset indicates the location of the fragment data referred to the start of the data block. Time to live contains a counter that indicates the life of the datagram. This prevents datagrams lost in the network from being transported continuously, producing unnecessary data traffic. Protocol specifies the protocol for the next higher layer, i.e. the application to which the datagram is to be delivered or from which it has been delivered. IP header checksum is used to check the IP header for possible transmission errors. IP source/destination address contains the IP address of the sender and receiver of the datagram. Options is used to adapt services to the higher protocols. For example, security requirements that must be met by the receive side can be specified here. Datagram contains the user data (1 to 65536 bytes).
ACTERNA
The World of IP
Page 2-10
IP Version 6 (IPv6) Frame Format IP Version 6 (IPv6) Frame Format

Bit 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Version Priority Flow label Payload length Next header IP source address (128-bit) Hop limit
IP destination address (128-bit)
Datagram (max. 65536 bytes)
Page 2-11
The rapid expansion of Internet access right down into the private user sphere rapidly led to address bottlenecks. The previous 32-bit addresses theoretically allow for more than 4 billion systems to be addressed. This is impossible in practice, since private local networks always occupy a larger or smaller range of addresses. Private networks can set up internal subnet addressing within their address range, which again means that a large proportion of IP addresses is lost. IP Version 6 provides for 128-bit addressing. This theoretically provides enough addresses for around 1500 computers per square meter of the earths surface. The IPv6 addresses are separated by colons instead of the dots used in IPv4 addresses. (Example of an IPv6 address: 2BA::66:834:459:AC3). The protocols and address formats of Version 4 and Version 6 are encountered in mixed systems. Version indicates the IP version. Priority indicates the priority of the datagram (lowest priority packets are rejected first if there is a router overload). Flow label is used by the data source to indicate data packets that require special handling with a specific quality of service. Next header indicates the protocol for the next higher layer. Hop limit represents a routimg counter. 1 hop corresponds to the passage of the packet through exactly one network node. The hop limit thus represents the life time of the packet.
ACTERNA
The World of IP
Page 2-11
Bridge //Switch //Router Bridge Switch Router

The links in the networks give the LAN / WAN its structure
Bridge/Switch Router
B/S R
for linking LAN segments for linking sub-networks
B/S
LAN segment LAN segment
WAN
Subnet
B/S
LAN segment
Subnet
LAN segment
Page 2-12
Coupling elements such as bridges, switches and routers connect and disconnect networks. They allow controlled access between network segments. Only that which should and may be transmitted to the other side is actually transmitted. Bridges and switches operate in the MAC layer. They generally connect LAN segments that are alike (Ethernet to Ethernet, Token Ring to Token Ring). Routers operate in the network layer (layer 3). This layer is independent of the technology used for transmission. Thus, routers are also used to link different LAN technologies and to couple LANs together via WANs (Ethernet to ISDN).
ACTERNA
The World of IP
Page 2-12
The Task of the Router The Task of the Router

How does a packet get from HERE to THERE through the many networks?
HERE Network 1 Decides on: Static routing Default routing Dynamic routing Network 3 R3 R1 Network 2 B
R2 Network 4 B THERE
Alternative routes
Page 2-13
Routing task: When two computers in different networks need to communicate with each other, the router must know the way to the receiver. It determines that the receiver is in another network from the address and subnet mask. It takes the routing information from its routing table. This routing table may only contain one active path to the destination, even if several paths to the receiver are possible. There are three possibilities: Static routing: Default routing: Path selection from fixed entries in the table. Path selection from one fixed entry in the table.
Dynamic routing: Automatic updating of the routing tables. Routing table R1 All traffic to Network 1: All traffic to Network 2: All traffic to Network 3: All traffic to Network 4: Direct via Port 1 Direct via Port 2 Direct via Port 3 To R2 via Port 2 with network metric = x or to R3 via Port 3 with network metric = y
(Network metric: cost factor, hops, speed, bit error rate, delay, throughput).
ACTERNA
The World of IP
Page 2-13
When Does a Router Route? When Does a Router Route?

1.1.1.5 A 1.1.2.5 F 1.1.3.5 G
Router 1
IP: 1.1.1.1 IP: 1.1.2.1 MAC: B MAC: C
Router 2
IP: 1.1.2.2 IP: 1.1.3.1 MAC: D MAC: E
1. 2. 3. 4.
1.1.1.5 1.1.2.5 A C 1.1.1.5 1.1.2.5 A B 1.1.1.5 1.1.3.5 A B 1.1.1.5 1.1.1.6 A B
X
1.1.1.5 1.1.2.5 C F 1.1.1.5 1.1.3.5 C D
X
1.1.1.5 1.1.3.5 E G
X
From MAC To MAC
Example: From IP To IP
IP address: Network Host 1.1.1.5

Page 2-14
The router only routes

if a data frame in the MAC layer is addressed to its MAC address and it can access the IP destination network.
Routers only use the network address in the Internet address of the destination system for routing through to the destination network.
Examples: 1. Not routed, since the MAC destination address is not the same as the router address. 2. Routed. The MAC destination address is B and the IP destination network can be reached from this router. 3. Routed by router 1 and router 2, as the correct MAC destination address is entered for both and the destination network (1.1.3.x) can be reached through both routers. 4. Not routed. The MAC destination address is that of router 1, but no routing is required to reach the specified IP destination network (1.1.1.x). The data frame is already in the destination network.
ACTERNA
The World of IP
Page 2-14
Layer 4 -- The Transport Layer Layer 4 The Transport Layer

Protected or simple data transport
The IP architecture offers two basic transport methods in the transport layer TCP - Transmission Control Protocol supports connection-oriented and protected data transport UDP - User Datagram Protocol supports connectionless and unprotected data transport
Page 2-15
Transmission Control Protocol: TCP is suitable for use in packet-switched networks or conglomerations of such networks. It provides virtual connection services for consecutive, protected transmission of user data. TCP is the basis for a protected inter-process communication mechanism above the unprotected packet switching layer (layer 3). TCP builds directly on IP and allows data loss detection, automatic retransmission of data packets, and duplicate packet detection. User Datagram Protocol: UDP provides application programs with a defined procedure that allows data (socalled datagrams) to be transmitted between communication partners with minimal protocol mechanisms. UPD sits directly on the Internet layer (IP, layer 3). It does not guarantee end-to-end control, unlike TCP. Hence it does not guarantee delivery of the datagram to the receiver, nor does it detect duplicate datagrams or ensure transmission of the data in the correct order.
ACTERNA
The World of IP
Page 2-15
One address (port) for the in-box

Layer 4 Task Port#1 File Port#2 E-mail Port#3 WWW Port#4
Layer 3 Layer 2 Layer 1
IP address MAC address Medium
Transport
Page 2-16
The network transports and routes the data frames from one user to another using the specified addresses (MAC/IP). The data also have to be passed on to the various applications within the user terminal (PC). A port is assigned to each application for this. The port is the address of the application in the user terminal. The functions from the medium to the application port layer have transportoriented tasks. The tasks are purely application-oriented from the port layer onwards. The individual functions are assigned to layers. The IP address function is in layer 3. The port address is in layer 4.
ACTERNA
The World of IP
Page 2-16
Application
The Port, the Application Address The Port, the Application Address
TCP Transmission Process TCP Transmission Process
Data TCP IP DLC
Data TCP IP DLC
Data
TCP SP/DP IP SA/DA MAC SA/DA
Page 2-17
Sequence of transmission process: A higher protocol transfers a data stream for transmission to the local TCP. The local TCP fragments the data stream into data segments and provides the following services: Setting up a full duplex connection Time monitoring of the transmitted data Numbering the data Flow control of all transmitted data Error checking
TCP transfers the data to IP. IP performs its services (formation of a datagram, fragmentation, etc.) and passes the data on to the data link layer. The data link layer packs the datagram into its data frames and transmits it to the destination computer. The received datagram is checked for errors, the data frame removed and the datagram transferred to IP by the data link layer of the destination computer. The local IP performs all the error check and reassembly procedures and passes the data to TCP. TCP performs its services, joins the received data segments back together to form the original data stream and passes this on to the higher protocol. DLC = Data Link Control SP/DP = Source Port / Destination Port SA/DA = Source Address / Destination Address
ACTERNA
The World of IP
Page 2-17
TCP/UDP Multiplex Mechanism TCP/UDP Multiplex Mechanism
Process #1
Process #2
Process #1
Process #2
TCP IP DLC TCP IP TCP IP TCP IP TCP IP IP
TCP
DLC
Page 2-18
TCP/UDP multiplex mechanism The multiplex mechanism provides the foundation for the coexistence of a number of higher protocols and processes within a single computer, as well as for the use of TCP/UDP by a number of processes within a single higher protocol. TCP/UDP assigns so-called port numbers in every computer to identify the various data streams. The entire exchange of data between the user processes and TCP/UDP takes place via these port numbers. Port numbers are assigned dynamically to the user processes and are random. Fixed port numbers are assigned to particular, frequently used processes; these are published as Assigned Numbers. The port number and the Internet address of the host are combined into a so-called socket to uniquely identify a TCP/UDP application process. A socket pair uniquely identifies a pair of processes that are communicating with each other.
ACTERNA
The World of IP
Page 2-18
TCP-A
TCP Connection TCP Connection

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. SYNC-SEQ = 0x100 ACK 0x101; SYNC-SEQ = 0x300
TCP-B
Connection setup
Page 2-19
ACK = 0x301 DATA-SEQ = 0x101 ACK-SEQ = 0x151 DATA-SEQ = 0x301 ACK-SEQ = 0x371 DATA-SEQ = 0x301 ACK-SEQ = 0x371 FIN-SEQ = 0x151 FIN-SEQ = 0x371; ACK = 0x152
ACK = 0x372
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.
Instance TCP-A opens the connection by synchronizing the sequence numbers. TCP-B acknowledges the connection and transmits the synchronization request for its sequence number. TCP-A acknowledges synchronization of the sequence number. TCP-A transmits a data packet of 0x50 bytes of data to TCP-B. TCP-B acknowledges receipt of 0x50 bytes of data. TCP-B transmits a data packet of 0x70 bytes of data to TCP-A. TCP-A acknowledges receipt of 0x70 bytes of data, but the block was damaged. TCP-B repeats transmission of the data packet to TCP-A. TCP-A acknowledges receipt of 0x70 bytes of data. Instance TCP-A terminates the connection. Instance TCP-B acknowledges clear down of the connection TCP-A acknowledges.
ACTERNA
The World of IP
Page 2-19
Connection clear-down
Data exchange
TCP Frame Format TCP Frame Format

1 2 bytes 2 bytes 4 bytes 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Source port Destination port Sequence number
4 bytes 2 bytes 2 bytes 2 bytes 2 bytes n bytes Header length
Acknowledgement number Reserved Window size Checksum Urgent pointer Options URG ACK PSH RST SYN FIN
Data
Page 2-20
The TCP packets are multiplexed via the source port and destination port. The sequence number indicates the sequence number of the current data packet. The received packets are acknowledged with the acknowledgement number. Header length specifies the further length of the header in 32-bit words. The control field contains 6 flags: FIN, SYN, RST, PSH, ACK and URG. These flags are used for acknowledgement, pointers, immediate transmission, resetting of a connection, sequence number synchronization and indication of the end of a connection. The window size field provides flow control by matching speed to the receiver capabilities. The header data checksum is checked via the checksum field. The urgent pointer can be ised to control priorities for important and urgent data.
ACTERNA
The World of IP
Page 2-20
UDP Frame Format UDP Frame Format
1 2 bytes 2 bytes 2 bytes 2 bytes
10
11
12
13
14
15
16
Source port Destination port Length of UDP Checksum of UDP
UDP Data
Page 2-21
The User Datagram Protocol (UDP) is a transport protocol (layer 4) in the OSI reference model and supports connectionless data exchange between computers. UDP was developed to allow application processes to transmit datagrams directly, so as to meet the requirements for transaction-oriented traffic. UDP builds directly on the subordinate IP. UDP does not guarantee that the datagram will be delivered top the destination user, and no provisions are made to prevent packet duplication or incorrect ordering of packets. UDP is described in RFC 768 along with other things.
ACTERNA
The World of IP
Page 2-21
How does acterna.com = 141.169.? How does acterna.com = 141.169.?

The domain name system (DNS), a service in the network
Internet
Top-level NS Primary NS DNS query LAN Local NS Forwarder NS
Page 2-22
Every computer in a TCP/IP network must have its own IP address. Use of a network can be greatly simplified for users if logical names can be entered instead of the complicated Internet addresses. This service is provided by the Domain Name Service (DNS). DNS was introduced in order to replace the central host table previously made available as hosts.txt by the NIC, with a distributed database application that would simplify maintenance of host and domain names. The result is that normally only the information about the network has to be loaded onto the local computer from the central file. The data about the hosts is recruited from the DNS. The DNS operates using an incremental hierarchy method. A query is initially processed by the local name server. This has access to all the address names in its area. If the query is outside this area, I.e. the name server does not have a corresponding entry, the query is referred to the server for the next higher hierarchy level. This server has access to all the address name assignments of its area, which is larger. The search procedure is repeated until either the root server is reached or a cross reference to another server is found. A positive search result leads to an entry in the caches of all the hierarchy levels below the server delivering the successful search result.
ACTERNA
The World of IP
Page 2-22
Chapter 3 Chapter 3
From GSM to GPRS Features and Basic Processes
The only constant thing is change

(Arthur Schopenhauer)
Page 3-1
Notes:
ACTERNA
Features and Basic Processes
Page 3-1
GSM Network Structure GSM Network Structure
PSTN
C D B
VLR
HLR AUC EIR
MSC A BSC Abis BTS
Page 3-2
A GSM network consists of the following components: Mobile terminal (MT) Base station subsystem (BSS) with controller (BSC) and transceiver (BTS) GSM switching center (MSC, mobile switching center)
In addition to this, it includes various network databases that provide user-specific information: Home Location Register (HLR) Visitor Location Register (VLR) Equipment Identification Register (EIR)
All these network components enable the GSM to provide circuit switched services.
ACTERNA
Page 3-2
BSS --Aspects BSS Aspects
GSM
BTS BTS BSC BSC
+?
GPRS
BTS BTS BSC BSC
BSS
BSS
Page 3-3
Various changes are required in the base station subsystem (BSS) to allow GPRS to be integrated into the existing GSM network. GPRS uses the same modulation, the same frequency band, the same method of frequency hopping and the same TDMA frame structure as conventional GSM. For this reason, the new Packet Data Channels (PDCH) are similar to the full-rate traffic channels (TCH) used for channel-switched speech services as far as the fundamental physical layer is concerned. GPRS can thus be considered as a GSM supplementary service. This means that the Base Transceiver Station (BTS), Base Station Controller (BSC) and the corresponding transmission paths already known from GSM can continue to be used, with slight modifications. In other words, complete GPRS coverage is guaranteed from day one onwards. The hardware changes are basically restricted to providing additional frequencies and paths to take account of increasing traffic. These expansions do not therefore need to be installed immediately, but can be added as and when required. The GPRS protocols defined in the GPRS specifications must be supported by the BSS. This requires an (extensive) software update.
ACTERNA
Page 3-3
GPRS Network Extensions GPRS Network Extensions
PSTN
C
Circuit switching
HLR AUC EIR
PSDN
Packet switching
Internet
Other operators
MSC A BSC
D B
VLR
GGSN
GGSN
SGSN
Abis BTS SGSN
Page 3-4
Expanding the existing GSM network for broadband applications based on packetswitched data requires implementation of certain packet-switched infrastructures within the GSM network to allow end-to-end packet transmission (GPRS). The following network elements have been introduced for this purpose: Serving GPRS Support Node (SGSN) and Gateway GPRS Support Node (GGSN)
The SGSN and GGSN network elements are needed exclusively for the packetswitched GPRS data services. The circuit-switched connections use the old GSM network architecture as before.
ACTERNA
Page 3-4
GPRS Network Structure GPRS Network Structure

EIR VLR HLR
MSC MSC BSS BSS Gb
PSTN
Gi SGSN SGSN Gn Gn GGSN GGSN IP
Backbone
SGSN SGSN
Gn
Gn GGSN GGSN
Gi X.25
Page 3-5
GPRS basiert im Gegensatz zur bisherigen leitungsorientierten GSM-bertragung auf einer paketorientierten bermittlung, d.h. der Netzbetreiber schaltet fr eine Verbindung keine Leitung vom Absender zum Empfnger durch. Vielmehr zerlegt GPRS den Datenstrom in Pakete, die einzeln und unabhngig voneinander den Weg zum Empfnger suchen. Erst im Endgert auf der Gegenseite werden die Pakete wieder in die richtige Reihenfolge gebracht. Die Betreiber mssen einzelne Teile des Netzwerkes neu aufbauen, bei einigen Teilen gengt eine leichte Modifizierung, andere wiederum bleiben unverndert. Keine nderung erfhrt das bestehende GSM-Kernnetz. Es bekommt lediglich eine Verbindung zum neu entstehenden GPRS-Netzwerk. Zentrale Elemente im IP-basierenden GPRS-Kernnetz sind der Serving GPRS Support Node (SGSN) und der Gateway GPRS Support Node (GGSN). Letzterer stellt die Schnittstelle zwischen der GPRS-Welt und anderen ffentlichen Datennetzen (z.B. Internet) dar. ber den GGSN laufen auch die Verbindungen zu anderen GPRS-Netzen, er ermglicht das Roaming mit anderen Netzen. Der SGSN bernimmt das Routing innerhalb des GPRS-Netzes. Die GSNs knnen durch ein internes (IP) Backbone-Netzwerk miteinander verbunden werden. Keine Erneuerung, aber eine nderung erfahren die Basisstationen. Je nach Bauart und Alter der BTS fllt die Anpassung unterschiedlich aus. Alle Funkschnittstellen bentigen ein Software Update, da bisher nur leitungsvermittelter Betrieb ber die Antennen lief. An manchen BTSen ist auch neue Hardware ntig. Die Kosten fr die Netzbetreiber sind erheblich. Ein SoftwareUpdate einer modernen BTS kostet um die 10.000 US Dollar. Pro Netz summieren sich die Umrstungskosten auf schtzungsweise 100 Millionen USDollar. Hinzu kommt der Aufwand fr den Aufbau des Backbones. Pro Netz drften 4 bis 8 Knoten (SGSN und GGSN) installiert werden.
ACTERNA Features and Basic Processes Page 3-5
Possible Bandwidths Possible Bandwidths

GSM
Type of connection Circuit-switched
HSCSD
Circuit-switched
GPRS
Packet-switched
Theoretical maximum data rate
14.4 kbit/s
115.2 kbit/s
(14.4 kbit/s x 8)
171.2 kbit/s
Realistic practical data rate
9.6 kbit/s to 14.4 kbit/s
28.8 kbit/s
approx. 40 to 50 kbit/s
Speech transmission
yes D1 D2 E-Plus Viag Interkom
no
no D1 D2 E-Plus Viag Interkom
Providers in Germany
D2 E-Plus
Page 3-6
Notes:
ACTERNA
Page 3-6
Was bestimmt die Bandbreite ? Was bestimmt die Bandbreite ?

GPRS Bandbreite = f(Kodierungsart und Anzahl der Timeslots)
Kodierungsart
CS 1 CS 2 CS 3 CS 4
9,05 13,4 15,6 21,4 1 TS
18,1 26,8 31,2 42,8 2 TS
27,2 40,2 46,8 64,2 3 TS
36,2 53,6 62,4 85,6 4 TS
45,3 67,0 78,0 107,0 5 TS
54,3 80,4 93,6 128,4 6 TS
63,35 93,8 109,2 149,8 7 TS
72,4 107,2 124,8 171,2 8 TS Timeslotanzahl
GPRS Datenrate in kbit/s
Page 3-7
Die verfgbare Datenrate hngt von dem Kodierschema (CS 1 ... CS 4) und von der Anzahl der fr den Paketdatenverkehr reservierten Kanle (Zeitschlitze) ab. Maximal sind so bei Belegung aller Timeslots einer Frequenz 171,2 kbit/s erreichbar. Das Kodierschema CS 4 bietet aufgrund der geringsten Redundanz (aber auch der geringsten Toleranz gegenber mglichen bertragungsfehlern) die maximale Netto-Bitrate. In der Praxis sind bei GPRS zunchst maximale Datenraten bis rund 50 kbit/s zu erwarten, was vier GSM-Zeitschlitzen entspricht. Anfangs werden die Netzbetreiber in einer Zelle immer nur ein paar Zeitschlitze fr GPRS freigeben. Das begrenzt dort fr alle Datenbertrager die Kapazitt auf die Zahl dieser Zeitschlitze z.B. 13,4 kbit/s. Ein einziger Dauerbertrager kann die Kapazitt bereits merklich verringern.
ACTERNA
Page 3-7
Important GPRS System Requirements Important GPRS System Requirements

Several parallel GPRS sessions per subscriber Simultaneous usage of different services Bandwidth compatible with standard packet-data networks (X.25, Internet) Channel independence Flexible assignment of radio resources by network operator Dynamic usage-oriented assignment of radio resources Data transfer compatibility with existing networks and applications
Page 3-8
Users should be able to run different GPRS sessions simultaneously (e.g. several Internet browsers active at the same time, etc.) Where the terminal equipment allows, simultaneous use of GPRS and classic GSM services (speech or data) must be possible. To ensure an adequate link to existing packet-switched networks, GPRS must provide a bandwidth that is at least comparable to that of a Frame Relay traffic channel. The GPRS multiplex mechanisms must operate independently of the available physical channels. The same procedure should then be possible later on for broadband channels (e.g. for UMTS). The GPRS network operator should be able to use O&M to configure the radio resources according to his requirements and considerations. It should be possible to change this configuration at any time to adjust to a given situation. A mobile phone must be able to recognize continuous configuration changes and to adapt accordingly. Any current sessions must not be terminated incorrectly as a result. It should be possible to change dynamically the amount of resources reserved for GPRS to correspond with the current need for data transmission capacity. The unused resources are then available e.g. for circuit-switched GSM services. The type and method of dynamic resource management is determined by the operator and is not specified by ETSI. ETSI does, however, specify the procedures necessary for this. GPRS must be compatible with existing data networks and applications. Existing applications must still be usable in conjunction with GPRS with only very small changes or no changes at all.
ACTERNA
Page 3-8
Fundamental Concepts Fundamental Concepts

Strict separation of radio subsystem and network subsystem Different QoS levels Standard applications (IP, X.25, etc.) and SMS are supported Flexible charge billing Security functions and access protection GPRS attach PDP context Transparent transmission of user-user data Dynamic assignment of available radio resources 3 terminal equipment classes Cell change
Page 3-9
The separation into a radio subsystem and a network subsystem allows the network side to be reused for other radio technologies. Thus, the existing GPRS network can be used later for UMTS. Different quality of service (QoS) levels are available. These can be selected to suit the type of data to be transmitted. General modern applications based on standard protocols such as X.25 or TCP/IP are supported as well as the Short Message Service (SMS) familiar from GSM. The GPRS operator can determine and invoice billing data on the basis of volume, QoS and connection time. The protection functions correspond to those already implemented in GSM. Procedures for authentication and activating data ciphering on the radio side are implemented in the SGSN. The algorithms correspond to those used by GSM. If an MS wishes to use GPRS, it must first log on to the network. This process is called GPRS Attach. The MS is only able to transmit packet data, to receive SMS over PS and to receive a paging message from the SGSN after this has been done.
ACTERNA
Page 3-9
To be able to transmit and receive packet data, the MS must first activate a socalled Packet Data Protocol Context (PDP context). This logs the MS in to the appropriate GGSN and it can start interworking between the GPRS network and an external packet data network (e.g. IP network).
User-user data are transported transparently between the MS and the external data network via GPRS network nodes. The method used to achieve this is called encapsulation or tunneling. This transparent transmission method largely avoids the need to interpret external protocols on the path through the GPRS network and provides for simple expansion for future interworking.
Special GSM radio channels are defined for GPRS. Assignment of these channels is flexible. 1 to 8 timeslots can be assigned to a MS per TDMA frame. Individual timeslots can be shared across several MSs (shared timeslot). The assignment of uplink resources is made separately from the assignment of downlink resources. The radio channels can be assigned dynamically according to load conditions and the requirements of the GPRS or GSM service operators. Various types of coding provide for transmission rates on the radio side from about 9 kbit/s up to more than 150 kbit/s per user. (A new modulation method means that EGPRS offers even higher transmission rates).
GPRS distinguishes between 3 operating modes: Class A Mode: Simultaneous GPRS and GSM services Class B Mode: A class B terminal monitors the GPRS and GSM control channels simultaneously, but can only use one GPRS or GSM service exclusively at a given time. Class C Mode: GPRS service only (pure data terminals)
The MS can perform a change of cell by itself or it is requested to change to a particular cell by the BSS. The MS informs the network of a completed change of cell.
ACTERNA
Page 3-10
Reference Points and Interworking Reference Points and Interworking

R reference point Um Gi reference point PDNs PDNs (oder andere Netze) PDNs (oder andere Netze) (or other networks)
TE TE
MT MT
GPRS network 1
MS
Gp reference point
GPRS network 2
Page 3-11
Network interworking functions are found at reference points Gi and Gp. It is possible that several interfaces to different data networks exist at the Gi reference point. These networks are often the property of different operators and use different communications protocols. The GPRS network operator has the task of defining the various interfaces at the Gi reference point and to negotiate with the PDN operators. In any case, the GPRS network must be able to work together with IP networks that use the IP protocol in accordance with specification RFC 791. Generally, the delays in a mobile data network are longer than those in the fixed network. This must be taken into account by the interworking function to ensure that calls are not cleared down unnecessarily because of timeouts.
ACTERNA
Page 3-11
High Level Functions High Level Functions

Network Access Control Packet Routing and Transfer Mobility Management Logical Link Management Radio Resource Management Network Management
Page 3-12
The Network Access Control Function includes: Registration (saving the subscriber data in the HLR) Subscriber identification and authentication Admission control (determining and reserving appropriate resources for the requested QoS) Packet terminal adaptation (converting the packet data into a form that can be processed by the GPRS network, e.g. segmenting, compression) Charging data collection (saving all connection data necessary to determine the call charges). The Packet Routing and Transfer Function includes: Relay function (in a network node: Receiving and forwarding data) Routing function (which is the next network node in the transmission path?) Address translation and mapping function (translates an IP address into a GPRS address for accessing a MS) Encapsulation function Tunneling function Compression function (data compression for efficient utilization of resources) Ciphering function (data ciphering on the radio side) Domain name server function (conversion of logical IP names into IP addresses and vice versa)
ACTERNA
Page 3-12
Mobility Management functions have the task of tracking a particular MS as it moves through the cells of a network and from one network to another. Information regarding the current location of a MS is also secured by Mobility Management functions.
Logical Link Management functions have the task of setting up, monitoring and clearing down (after data transfer is complete) logical connections between the network and the mobile station.
Radio Resource Management functions look after the assignment and management of the available radio channels.
Network Management functions provide an O&M interface between the GPRS transmission network and the OMC.
ACTERNA
Page 3-13
Logical Architecture Logical Architecture

SMS-GMSC SMS-GMSC SMS-IWMSC SMS-IWMSC E Gd MSC/VLR MSC/VLR A R Um Gs D HLR HLR Gr Gc Gi C SM-SC SM-SC
TE TE PDN
TE TE
MT MT
BSS BSS
Gb
SGSN SGSN
Gn
GGSN GGSN
Gn
Gp
Gf
GGSN GGSN SGSN SGSN further PLMN
EIR EIR
Signaling Signaling and data
Page 3-14
Logical implementation of a GPRS system is based on the usual GSM structure. Only two new network nodes are added, the Gateway GPRS Support Node (GGSN) and the Serving GPRS Support Node (SGSN). It has also been necessary to define several new interfaces (Gb, Gc, Gd, Gf, Gi, Gn, Gp, Gr, Gs). These logical interfaces do not all have to be implemented in a real system. For example, the signaling between SGSN and VLR can also be handled via the Gb and A interfaces instead of Gs.
Internal addressing of the GGSN and SGSN network nodes Both nodes are addressed for network-internal communications using IP addresses via the GPRS backbone network. The IP addresses of all GPRS nodes form a private address area that is not accessible from public networks. The GGSN and SGSN normally can be addressed using logical DNS names as well.
ACTERNA
Page 3-14
GGSN tasks The GGSN forms the gateway between the GPRS network and the external (public) packet data network. The GGSN contains all the routing information for every mobile terminal registered using the GPRS Attach procedure. The GGSN has access to the HLR for querying subscriber and location data. Based on the available routing information, the PDN protocol frames are tunneled from the GGSN to the SGSN.
SGSN tasks The SGSN is the node directly serving the mobile station and which performs all necessary protocol adjustments. For example, the very long IP packets are segmented here into smaller data packets that are then reassembled in the correct order on the MS side. During the GPRS Attach procedure, a so-called Mobility Management context is created in the SGSN. This data structure contains all the necessary information regarding the mobility and the security functions of a MS. When the MS logs on, a PDP (Packet Data Protocol) context is set up in the SGSN. This contains all the routing information and is used together with the GGSN for routing through to the mobile station. The SGSN can communicate directly with MSC/VLR via the optional Gs interface in order to update location information or to receive paging requests.
Both GPRS support nodes (GSN), i.e. GGSN and SGSN, can be combined physically into one unit, or they can be located in different physical system components that are linked by the Gn interface and perhaps an intervening IP network. Both nodes can process IP routing and can be networked via IP routers. If a SGSN is connected to a GGSN located in a different mobile radio network, the connection is made via the Gp interface. Gp has the same function as the Gn interface with the addition of security functions (firewall, etc.). The scope of the security functions is determined by mutual agreement between the two mobile radio network operators.
ACTERNA
Page 3-15
GPRS Backbone Networks GPRS Backbone Networks

Packet data network
Gi Gp
Inter-PLMN backbone Gp
Gi
GGSN GGSN
BG BG
BG BG
GGSN GGSN
Intra-PLMN backbone
Intra-PLMN backbone
SGSN SGSN PLMN A PLMN A
SGSN SGSN
SGSN SGSN PLMN B PLMN B
Page 3-16
There are two types of backbone network within a global GPRS network structure: 1. 2. Intra-PLMN backbone network Inter-PLMN backbone network
Every intra-PLMN backbone network corresponds to a purely private IP network that serves only to transmit GPRS data and GPRS signaling packets. The intraPLMN backbone network connects the SGSN and GGSN at the Gn reference point within a specific PLMN. Two intra-PLMN backbone networks are connected together via an inter-PLMN backbone network. The network gateway takes the form of a border gateway (BG), which should include an appropriate level of security functions. Definition of the BG is not part of the GPRS specification and is left up to the network operators. A packet data network (e.g. public Internet) or leased lines are suitable as an inter-PLMN backbone network.
ACTERNA
Page 3-16
Mobility Management Mobility Management

Status changes in the MS
IDLE
GPRS attach GPRS detach
READY
Timeout
PDU transmission
STANDBY
STD for MS mobility management
Page 3-17
Mobility Management services are realized through communications between the mobile terminal and the SGSN or MSC/VLR. They allow the mobile terminal access to GPRS-GSM services and give it mobility (location management). The Mobility Management process (MM) can take on the following states: IDLE READY STANDBY A range of different activities is possible in each of these states. Special information is required in the mobile terminal and in the SGSN to perform the MM functions. This set of information is also called a Mobility Management context (MM context).
IDLE state
In this state, the mobile terminal is not connected to the MM function. Both the mobile terminal and the SGSN do not have any valid context information regarding routing or position. There are no subscriber-related MM procedures active. The GPRS terminal is considered as bein unreachable in this state. The mobile terminal can only receive point-to-multipoint (broadcast) messages in this state. An ATTACH procedure is required before MM contextsare set up in the MS and SGSN.
ACTERNA
Page 3-17
Mobility Management Mobility Management

Status changes in the SGSN
IDLE
GPRS attach GPRS detach Cancel location
Implicit detach Cancel location
READY
Timeout
PDU reception
STANDBY
STD for SGSN mobility management
Page 3-18
STANDBY state This state is assumed after a successful ATTACH procedure. The MS and SGSN now each have a MM context for the IMSI of the mobile terminal. PAGING messages for PTP and GSM-CS (circuit-switched) can be received as well as point-to-multipoint (PTM) data. In this state, the mobile terminal selects a local GPRS routing area (RA) and a GPRS cell for future communications. The mobile terminal also executes various MM procedures in order to inform the SGSN about any routing area changes. Cell changes within the same RA are not reported. For this reason, the SGSN only keeps the GPRS routing area identifier (RAI) in its MM context in the STANDBY state. A mobile terminal can initiate activation or deactivation of a PTP context from this state. PTP context activation is required before a data communication. If an activated PTP context is available, PTP packets can also be received in STANDBY state. In this case, the SGSN sends a PAGING request message in the routing area where the MS is located. This, however, depends on the status of the Paging Proceed Flag (PPF). If the mobile terminal responds to the paging message, the transition from STANDBY to READY starts to take place in the MS and SGSN. READY state In this state, the SGSN context also contains the cell information for the mobile terminal. This takes place from the terminal with the aid of MM procedures. Cell selection takes place locally through the MS but can also be controlled from the network (SGSN). In the READY state, all types of data packets can be received and transmitted. The cell identifier is contained in the BSSGP header of the data packets from the mobile terminal. A detach procedure leads from the READY state to the IDLE state. If a READY state monitoring timeout occurs, the status changes from READY to STANDBY.
ACTERNA
Page 3-18
MM: Interactions Between SGSN and MSC/VLR MM: Interactions Between SGSN and MSC/VLR
Messages between SGSN and VLR are exchanged over the Gs interface to combine the following GSM and GPRS processes using resources as efficiently as possible:
IMSI attach and detach via SGSN Coordination between GSM and GPRS location update Paging a GSM connection via the SGSN Identification Alert procedures for non-packet services MM information procedures
Page 3-19
Signaling between the SGSN and MSC/VLR can be transmitted via the Gs interface, with the aim of setting up an association between the two network components with regard to a particular MS. Once the association is made, information can be exchanged via it directly. For example, an MS that already has a GPRS connection (i.e. is GPRS attached) can transmit the messages over the radio channels assigned for GPRS for setting up an additional, parallel GSM MOC connection. The SGSN then communicates directly with the MSC. Also, combined location updates can be sent from the MS to the SGSN. The SGSN then passes the location update information on to the VLR. The association between the MSC/VLR and SGSN is updated as soon as the MS changes the VLR area or the SGSN.
ACTERNA
Page 3-19
MM: Interactions Between SGSN and MSC/VLR MM: Interactions Between SGSN and MSC/VLR
Example: CS paging
BSS BSS
SGSN SGSN
MSC/VLR MSC/VLR
Paging Paging Paging
Paging response (SABM)
Paging response
(SCCP connection request)
Page 3-20
Whenever a mobile station is IMSI and GPRS attached, the mobile radio network can transmit a CS Paging message (CS = circuit switched) via the SGSN. If the MS is in MM STANDBY state, the paging message will be broadcast in the routing area. In the READY state, the paging message is transmitted directly to the cell in which the MS is located.
ACTERNA
Page 3-20
MM: Attach Procedure MM: Attach Procedure

MS MS
Attach request Identification request Identity request Identity response Authentication IMEI check Update location Cancel location Cancel location acknowledge Identification response Authentication IMEI check
SGSN (new) SGSN (new)
SGSN (old) SGSN (old)
EIR EIR
HLR HLR
Update location acknowledge Attach accept
Attach complete
Page 3-21
The ATTACH procedure places the mobile terminal and the SGSN in the READY state. It is therefore the prerequisite for bi-directional data communications. There are two types: GPRS attach for activation of GPRS services, and IMSI attach for activating the standard GSM services. Depending on the network operating mode, a mobile terminal can be connected simultaneously as a GPRS and as a GSM (IMSI) MS. This is known as a combined GPRS/IMSI attach and a combined routing area / location area update (RA/LA) procedure. This is only possible for Mode 1 networks. The mobile terminal performs a non-combined GPRS attach procedure and an RA update procedure in network operating Mode 2 and Mode 3. The mobile terminal indicates its identity (P-TMSI or IMSI) and the attach type (GPRS or IMSI) to the SGSN during the attach procedure. If the MS does not have a valid P-TMSI, it uses its IMSI to log in. The MS logs in on the radio link control / media access control (RLC/MAC) layer with a temporary logical link identifier (TLLI) if such has already been assigned. Otherwise, it uses a random TLLI. This TLLI is used to address the MS until a new P-TMSI is assigned (the TLLI is derived directly from the P-TMSI). After the attach procedure has been completed, the mobile terminal is in the READY state and the necessary MM contexts have been set up in both the mobile terminal and the SGSN. The MS can then activate PDP contexts and begin packet-switched communications. The SGSN address of the mobile terminal that is logged in is stored in the HLR database. The diagram above shows a simplified GPRS attach process.
ACTERNA
Page 3-21
The attach procedure is initiated by the MS with the Attach Request message and completed with Attach Accept. Various activities take place between these messages, i.e. : Identification of the mobile terminal Authentication Updating of the stored location information between the SGSN, HLR and VLR.
The most important parameters in the Attach Request message are: IMSI P-TMSI Classmark if no valid P-TMSI is available Packet Temporary Mobile Subscriber Identity additionally defines the multi-timeslot properties and the available ciphering algorithms of the MS Ciphering Key Sequence Number GPRS, IMSI, combined GPRS/IMSI Facility for discontinuous reception for the MS
CKSN Attach type DRX Old P-TMSI signature
The P-TMSI signature represents a further aspect for checking and security of the P-TMSI. If the specified signature does not match the one stored in the SGSN, the SGSN can activate further security functions to authenticate the mobile terminal.
ACTERNA
Page 3-22
Trace Example Trace Example
08154711
Page 3-23
The trace example shows the progress of a GPRS attach process with an existing IMSI attach.
ACTERNA
Page 3-23
MM: Detach Procedure MM: Detach Procedure

SGSN SGSN
Detach request Delete PDP context request
GGSN GGSN
MSC/VLR MSC/VLR
Delete PDP context response
GPRS detach indication
Detach accept
Page 3-24
The diagram shows a detach procedure initiated by the network as an example. The SGSN uses Detach Request to inform the MS that it will be disconnected. The Detach Type parameter can be used to request the MS to perform a new attach procedure immediately after the detach procedure to reactivate the previously deactivated PDP context. The Delete PDP Context Request message requests the SGSN to deactivate the active PDP context for this particular mobile station. The GGSN acknowledges execution of this request with Delete PDP Context Response. If the MS was IMSI as well as GPRS attached, the SGSN sends a GPRS Detach Indication message to the VLR. The VLR then deletes the association between the VLR and SGSN for this MS and performs any future paging or location update procedures without the participation of the SGSN.
ACTERNA
Page 3-24
MM: Security Functions MM: Security Functions

Subscriber Authentication Procedure
SGSN SGSN
Send authentication info
HLR
Send authentication info acknowledgement
Authentication and ciphering request
Authentication and ciphering response
Page 3-25
The Mobility Management security functions provide security with regard to: Unauthorized use of GPRS (checking of subscriber and service requested) Maintenance of user confidentiality (P-TMSI and signaling data ciphering) Maintenance of user data confidentiality (user data ciphering) The procedures already defined for GSM are used for authentication of the subscriber, except that these procedures are performed in the SGSN for GPRS. The authentication procedure allows the subscriber identity to be determined or a ciphering algorithm to be set or both together. The parameters necessary for authentication are usually stored in the SGSN as long as the MS is within its location area. If the SGSN does not have the authentication parameters for a particular MS, it initially requests them from the HLR with a Send Authentication Info message. The HLR returns the information by means of the Send Authentication Info Acknowledgement message. The Authentication and Ciphering Request message contains the parameters RAND and CKSN as well as the ciphering algorithm to be used. The MS transmits the SRES parameter to the SGSN with the Authentication and Ciphering Response message.
ACTERNA
Page 3-25

User Identity Confidentiality
SGSN SGSN
P-TMSI reallocation command
P-TMSI reallocation complete
Page 3-26
A GPRS user is identified by a so-called Temporary Logical Link Identifier (TLLI). The relationship between TLLI and IMSI is only known to the mobile station and the SGSN. The TLLI is derived from the P-TMSI, which is assigned temporarily to the MS by the SGSN. The SGSN may change the P-TMSI at any time and reallocate it to the MS. The diagram shows a P-TMSI reallocation procedure.
ACTERNA
Page 3-26

Identity Check Procedure
SGSN SGSN EIR
Identity request
Identity response
Check IMEI
Check IMEI acknowledge
Page 3-27
GPRS uses the procedures already defined for GSM to check identity. The Identity Request message contains an Identity Type parameter as a message element. This indicates the item that is to be checked (IMSI, TMSI or IMEI). The example above shows the sequence for an IMEI check procedure. After receiving the IMEI, the SGSN starts a corresponding query to the EIR and receives positive or negative acknowledgement in the form of the Check IMEI Acknowledgement message.
ACTERNA
Page 3-27
MM: Location Management MM: Location Management

Intra-SGSN Routing Area Update Procedure
SGSN SGSN
Routing area update request
Security functions
Routing area update accept
Routing area update complete
Page 3-28
Location Management supplies functions for: Cell and network selection Routing area update (in STANDBY and READY states of an MS) Cell update (in READY state of an MS) A routing area is a subset of a location area. A LA is made up from at least one RA. An RA is managed exclusively by a single SGSN. The diagram shows an intra-SGSN routing area update procedure. The MS sends a Routing Area Update Request message to the SGSN (the BSS adds the old RAI, the old P-TMSI and the update type to this). If required, security procedures are then performed (ciphering, authentication, TMSI reallocation). When all the prerequisites for an RA change are met (authorizations, no errors in the security procedure, etc.), the SGSN updates the MM context for the affected MS and responds with Routing Area Update Accept. Once the MS is assigned a new TMSI, it responds with Routing Area Update Complete. If an RA update procedure fails several times or the SGSN responds with Routing Area Update Reject, the MS switches to IDLE state.
ACTERNA
Page 3-28
MM: Location Management (Inter-SGSN RA Update) MM: Location Management (Inter-SGSN RA Update)
MS MS 1
RA update request SGSN context request
SGSN (new) SGSN (new)
SGSN (old) SGSN (old)
GGSN GGSN
HLR HLR
2 SGSN context response

Security functions Security functions
3 4
SGSN context acknowledge Forward packets Update PDP context request Update PDP context response Update location Cancel location Cancel location acknowledge Insert subscriber data Insert subscriber data acknowledge Update location acknowledge
RA update accept RA update complete
Page 3-29
The GGSN and HLR are also involved in an inter-SGSN RA update. The MS sends a RA Update Request to the SGSN. The parameters are the old RAI, old P-TMSI and update type. The new SGSN requests the MM context and PDP context from the old SGSN with a SGSN Context Request message. The parameters in this message are the old RAI, TLLI, old P-TMSI and the address of the new SGSN. The MM context and PDP context are indicated to the new SGSN in the SGSN Context Response message. The new SGSN responds with SGSN Context Acknowledge to indicate that it is ready to receive the data packets for this active context. If the old SGSN still has data packets stored that have not been transmitted, it now passes these on to the new SGSN. Any data packets sent in Acknowledged Mode to the MS which have not yet been acknowledged will also be sent to the new SGSN. The GGSN is asked to update its PDP context with the Update PDP Context Request message. The GGSN responds with Update PDP Context Response. The HLR is informed of the SGSN change by the Update Location message. The HLR then uses a Cancel Location Message to request the old SGSN to delete its MM context and PDP context. The HLR transfers the subscriber data to the new SGSN in the Insert Subscriber Data message. The MS receives a new P-TMSI with the RA Update Accept message and acknowledges this with RA Update Complete.
ACTERNA Features and Basic Processes Page 3-29
MM: Classmark Handling MM: Classmark Handling
Mobile station classmark
Radio access classmark
MS network capability
Page 3-30
To make it easier to introduce new radio access technologies in the future (CDMA in conjunction with UMTS, etc.), a distinction is made between the Radio Access Classmark and the MS Network Capability when specifying the performance features and physical facilities of an MS. Together they are referred to as the MS Classmark. Mobility management messages are used to transmit the classmark information from the MS to the network, where it is stored. As long as the MS is attached, no further classmark information needs to be transmitted over the radio interface. The Radio Access Classmark contains all the information about the radio capabilities of an MS, e.g. multislot capability, frequencies, power class, etc.). The MS Network Capability describes all other radio-dependent characteristics of an MS, e.g. available ciphering algorithms, coding, etc.
ACTERNA
Page 3-30
Radio Resource Functions Radio Resource Functions

Cell selection and reselection Discontinuous reception (DRX) Radio resource management Paging
Page 3-31
Cell Selection and Reselection An MS can only be attached to one cell (serving cell) at any given time. The cell selection and reselection procedure allows the MS to select the cell in which the desired service is also offered; for example, an MS requiring GPRS will not change to a cell that may be receivable with a good BCCH level but which does not provide GPRS. Cell selection and reselection can be performed by the MS alone or it can be controlled by the network. Discontinuous Reception A GPRS MS can itself determine whether or not to use DRX. If DRX is used, the MS indicates the DRX parameters to the network within the attach procedure. The network determines the correct paging group from these parameters together with the IMSI. Radio Resource Management has the following tasks: Occupation and release of physical resources (timeslots) Monitoring of GPRS channels to determine usage or overload Resolving overload situations Indicating the GPRS channel configurations in the form of broadcast messages Paging The paging procedure transfers the MS from the MM IDLE state to the MM READY state. In this state, the SGSN can transmit data packets to the corresponding MS.
ACTERNA
Page 3-31
Radio Resource Functions Radio Resource Functions

Paging
BSS BSS
SGSN SGSN
1 2 Paging request PDP PDU
GPRS paging request
Any LLC frame
4 Any LLC frame 5
Page 3-32
The SGSN receives a downlink PDP PDU (data packet) intended for a particular MS. The SGSN sends a paging request message with the parameters IMSI, PTMSI, area, channel needed, QoS, and DRX to the BSS. The IMSI is needed for the BSS to determine the correct paging group. The BSS sends a paging request message to every cell belonging to the specified routing area. , Every valid LLC frame received after a paging request message is interpreted as a response to the preceding paging request message. However, before an LLC frame can be transmitted by the MS, the necessary packet assignment procedures for occupying GPRS channels must be performed.
ACTERNA
Page 3-32
Packet Routing and Transfer Packet Routing and Transfer

Functions

Packet routing and transfer function Relay function Packet terminal adaptation function Encapsulation function
Page 3-33
The Packet Routing and Transfer function takes care of addressing and transfer of data packets between the MS (reference point R) and the external network (reference point Gi) the further GPRS network (reference point Gp) further MSs (reference point R) The maximum packet size is 1500 octets. Longer packets are segmented. Packet communications between the MS and SGSN are taken care of by the SNDCP protocol. The Relay function takes care of converting and adjusting (e.g. segmenting, compressing, buffering) the data packets between the input and output sides of a network node (e.g. SGSN). The Terminal Adaptation function takes care of converting the received data packets from terminal equipment conforming to the X.28, X.29, etc. standards. In GPRS, the data packets are transmitted transparently between the MS and the external network (e.g. Internet) over the mobile radio network. To facilitate routing, all packets are packed maybe several times on the path through the GPRS network and are unpacked at their destination. These mechanisms are part of the Encapsulation function. Different methods of data encapsulation are used on the transmission paths from SGSN to GGSN and from MS to SGSN.
ACTERNA
Page 3-33

PDP states
INACTIVE INACTIVE
PDP context deactivation or MM status change to MM-IDLE
PDP context activation
ACTIVE ACTIVE
Page 3-34
A point-to-point GPRS relationship contains one or more PDP addresses (PDP = Packet Data Protocol). These PDP addresses are described in the MS, SGSN and GGSN in each case by a PDP context. A PDP context (data set) contains all the necessary information about routing, QoS profile, etc. Every PDP context for a GPRS subscriber is associated with the subscribers MM context. Definition of Packet Data Protocol states INACTIVE: In this state, the PDP context does not contain any valid routing information. No transmission of data packets is possible. A subscriber location update does not lead to an update of the PDP context in this state, not even if the subscriber is GPRS or IMSI attached. ACTIVE: In this state, the MS, SGSN and GGSN all have an active PDP context. All the information necessary for routing, etc. is stored in this PDP context. Transmission of data packets is now possible between the MS and GGSN. The PDP ACTIVE state can only be assumed if the MM state is STANDBY or READY.
ACTERNA
Page 3-34

PDP context activation
SGSN SGSN
GGSN GGSN
Activate PDP context request 2 Security functions 3 Create PDP context request
4 Create PDP context response 5 Activate PDP context accept
Page 3-35
The diagram shows PDP context activation initiated by the MS. The MS sends an activation request to the SGSN. The parameters include NSAPI, TI, PDP type, and PDP address. Security functions (ciphering, etc.) are activated if required. , Once the SGSN has checked the activation request, it responds with the Create PDP Context Response message. Activation of the PDP context is acknowledged to the MS.
ACTERNA
Page 3-35

PDP context modification
GGSN GGSN
SGSN SGSN
Update PDP context request
Update PDP context response
Modify PDP context request
Modify PDP context accept
Page 3-36
If parameter changes (QoS profile, etc.) make it necessary to change an existing PDP context, this can take place as shown in the example here.
ACTERNA
Page 3-36

PDP context deactivation
GGSN GGSN
Delete PDP context request
SGSN SGSN
Delete PDP context response
Deactivate PDP context request
Deactivate PDP context accept
Page 3-37
This example shows deactivation of a PDP context initiated by the SGSN.
ACTERNA
Page 3-37
MM and PDP context for the SGSN
ACTERNA
Page 3-38
MM and PDP context for the MS
PDP context for the GGSN
ACTERNA
Page 3-39
Transmission Transmission
Layer model
Application IP/X.25 Relay SNDCP SNDCP LLC RLC RLC MAC GSM RF MAC GSM RF Relay BSSGP BSSGP Network service L1bis Network service L1bis IP L2 L1 IP L2 L1 LLC GTP UDP/ TCP UDP/ TCP GTP IP/X.25
MS
Um
BSS
Gb
SGSN
Gn
GGSN
Gi
Page 3-40
The GTP, LLC and RLC protocols provide for various transmission modes. The combination of GTP, LLC and RLC determines the QoS reliability class. GTP transmission modes: The GTP protocol provides two transmission modes for user data 1. Unacknowledged mode (UDP/IP) 2. Acknowledged mode (TCP/IP) Both transmission modes can be used simultaneously. LLC transmission modes: Similarly, two modes that can be used simultaneously are defined for the LLC layer. 1. Unacknowledged mode. Received LL PDUs are not acknowledged. 2. Acknowledged mode. Acknowledgement is performed in the LLC layer. If no acknowledgement is received, data packets are repeated after the timeout has elapsed. RLC transmission modes: Again, acknowledged and unacknowledged modes are provided here. Both modes can be used simultaneously.
ACTERNA
Page 3-40
LLC functions
Transmission of SNDCP protocol data units (SN-PDUs) between the SNDCP Transmission of LLC-PDUs between the MS and SGSN in protected or Point-to-multipoint transmission between the SGSN and several MSs Detection and regeneration of lost or damaged LLC-PDUs Flow control Ciphering of LLC-PDUs
unprotected operation and LLC layers
Page 3-41
Notes:
ACTERNA
Page 3-41
Sub-network dependent convergence
User data User data (Packet data protocol) (Packet data protocol) N-PDU NSAPI SNDCP SNDCP SNDCP header NSAPI + Control SAPI SAPI SAPI Data
Signaling Signaling
SMS SMS
LLC LLC
LLC header TLLI Control LLC information Data
RLC@MS or BSSGP@SGSN RLC@MS or BSSGP@SGSN

Multiplexing the network protocols
Page 3-42
The Subnetwork Dependent Convergence Protocol (SNDCP) is within the layer architecture between the network layer and the LLC layer in the MS and the SGSN. Different network layers (e.g. X.25, IP) are supported. The network layer protocols all access a common SNDCP protocol instance. The required multiplex formation takes place via the Network Layer Service Access Point Identifier (NSAPI). One of the main tasks of the SNDCP is compression and segmenting of the user data packets that are to be transmitted. The users above LLC (signaling, SMS, SNDCP) use different Service Access Point Identifiers (SAPI) to access the LLC layer services. The MS is addressed using the Temporary Logical Link Identifier on the RLC layer (in the MS) and the BSSGP layer (in the SGSN).
ACTERNA
Page 3-42
Packaging the Protocols at the Air Interface Packaging the Protocols at the Air Interface
User data block
40 bytes
Application
TCP/IP header
2 or 3 bytes
IP (or X.25) packet
Header
>= 6 bytes Maximum 1600 bytes
SNDCP
4 bytes
Header
181 Bit RLC/MAC header
Information frame
FCS
LLC
LLC frame
BCS
RLC/MAC
456 bits
Convolution coding
NB 1 114 bits
NB 2 114 bits
NB 3 114 bits
NB 4 114 bits
Air interface
Tail 3 bits
Data 57 bits
S 1 bit
Training sequence
S 1 bit
Data 57 bits
Tail 3 bits
Page 3-43
The application accessing a packet service via an MS is superimposed on the network layer, usually TCP/IP or X.25. The user data packets are mapped on to the LLC layer by the Subnetwork Dependent Convergence Protocol (SNDCP). The SNDCP can split IP or X.25 frames into individual segments and can multiplex network layer frames together that pertain to different applications or sessions, in order to transmit them over a virtual connection. SNDCP also allows data compression and data ciphering. The Logical Link Control layer (LLC) in GPRS takes on a similar function to LAPDm in GSM, I.e. it looks after flow control and error correction. LLC allows reliable transmission of user data between the mobile subscriber and the SGSN. The LLC layer also supports point-to-multipoint transmissions. The Radio Link Control layer (RLC) accepts the LLC data and provides a reliable locical connection between the MS and BSS. Access to the GPRS channels is controlled by the Media Access Control layer. As a large number of GPRS subscribers all make use of a pool of physical channels reserved for packet services at practically the same time, access control determines who can occupy which timeslot at which carrier frequency and at what time. The main tasks of the MAC layer are to multiplex the subscribers onto the common resources, collision detection and clearance, and adherence to a reservation strategy for the agreed quality of service. The MAC layer can assign an MS several physical channels at the same time to increase the data rate. The data from the MAC are then prepared for radio transmission by appropriate channel coding and then transmitted using the usual GSM normal bursts.
ACTERNA
Page 3-43
Identifiers Identifiers
IMSI
RAI
TEID
NSAPI P-TMSI GSN address
CI TLLI PDP address
Page 3-44
IMSI: Represents an internationally unique subscriber number. It consists of the mobile country code, mobile network code and mobile station identification number. P-TMSI: The Die Packet Temporary Mobile Subscriber Identity (P-TMSI) is assigned temporarily to an MS as long as the MS is in the location area of an SGSN. The P-TMSI is then transmitted instead of the IMSI in the messages on the radio interface. This measure makes it more difficult to track a mobile using its IMSI. NSAPI and TLLI: The Network Service Access Point Identifier (NSAPI) and Temporary Logical Link Identifier (TLLI) are used for routing on the network layer. The NSAPI identifies the PDP SAP in the MS. In the SGSN and GGSN, the NSAPI identifies a PDP context, which is linked to a PDP address. The TLLI uniquely identifies the logical link between MS and SGSN. Within a routing area, there is a one-to-one relationship between the TLLI and IMSI. This relationship between TLLI and IMSI is only known in the MS and the SGSN.
ACTERNA
Page 3-44
PDP address: A GPRS subscriber identified by an IMSI has one or more network addresses, called PDP addresses (PDP = Packet Data Protocol). The PDP addresses are assigned temporarily for the duration of the packet connection or can be assigned permanently. The PDP address format corresponds to the address format of the relevant network layer. When the MS is connected to an IP network (e.g. Internet), the PDP address is coded according to the IP Version 4 or Version 6 address format. The PDP address is set up by the PDP Context Activation procedure and is deleted by the PDP Context Deactivation procedure. TEID: The Tunnel Endpoint Identifier (TEID) is used by the GPRS tunneling protocol between the GSNs as a logical address that identifies a tunnel. RAI: The Routing Area Identity (RAI) is assigned by the network operator and defines a routing area. An RA consists of at least one cell, but usually several cells. The RAI is transmitted as a system broadcast message. An RA is a subset of a location area and is handled exclusively by a single SGSN. CI: The Cell Identity (CI) identifies a cell within a routing area or location area. GSN address: The GSN address is used to address the SGSN or GGSN via the internal backbone network. Both network nodes have at least one IPv4 address and an optional IPv6 address. As the IP addresses of the GSNs are only intended for communications via the internal GPRS network, and therefore form a private address area, the network nodes cannot be addressed from the public Internet using their GSN addresses.
ACTERNA
Page 3-45
Identifiers Identifiers
Use of NSAPI and TLLI
IP address A GPRS MS GPRS MS
IP application 1 (IP address A) IP application 2 (IP address B)
IP address B
NSAPI 1
NSAPI 2
SNDCP
IP network
LLC
TLLI
SGSN SGSN GGSN GGSN
Page 3-46
Multiplexing can be performed in the network layer using the NSAPI (Network Layer Service Access Point Identifier). This allows several parallel Internet transactions as shown in the above example.
ACTERNA
Page 3-46
Quality of Service Profile Quality of Service Profile

Every PDP context is assigned a particular QoS profile. The QoS profile describes the possible service quality for a particular connection. Every QoS profile consists of these 5 attributes:

precedence class delay class reliability class peak throughput class mean throughput class
Each attribute has a defined range of values, so a large number of different QoS profiles are possible for a given connection.
Page 3-47
Not all of the QoS profiles possible from the large number of combinations have to be available in a GPRS network. The attribute values and therefore the QoS profile can be negotiated between the MS and the network. The network should negotiate each parameter according to the available network resources. Precedence Class: The PC defines the priority of a connection (e.g. with regard to network problems or overloads: can packets be destroyed or not).
Delay Class: GSM 02.60 defines 4 delay classes. The network must basically provide sufficient transmission resources to guarantee the maximum delay within a delay class and cell. Delay class 4 corresponds to best effort.
ACTERNA
Page 3-47
Reliability Class: The RC describes the probability of a residual error rate for the following: Data loss Sequence errors Duplicated transmissions Transmission errors (bit errors) The Reliability Class specifies the requirements to the various protocol layers. The combination of GTP, LLC and RLC modes leads to a particular Reliability Class.
Signaling and SMS messages are transmitted with Reliability Class 3. Throughput Classes: The user data throughput is given in terms of Throughput Class units. In this way, the expected bandwidth is described in conjunction with a particular PDP context. A distinction is made between Peak Throughput and Mean Throughput classes. The peak throughput class defines the maximum peak throughput in kbit/s. There is, however, no guarantee that this data throughput will be achieved at any given time. The network can strongly limit the peak throughput during the negotiation procedure, even if sufficient resources are available. The mean throughput class defines the realistic mean value of data throughput for the remaining time of a specific active PDP context. Here, too, the network can strongly limit the mean throughput class during the negotiation. The minimum that can be negotiated is a Best Effort Mean Throughput Class.
ACTERNA
Page 3-48
Notes:
ACTERNA
Page 3-49
Quality of Service Negotiation Quality of Service Negotiation
SGSN SGSN
Activate PDP context request {QoS requested}
Activate PDP context accept {QoS negotiated}
Page 3-50
During the PDP context activation procedure already described, the MS communicates the desired quality of service profile using the QoS Requested parameter. The network indicates the actual QoS profile to the MS after validation in the form of the QoS Negotiated parameter within the Accept message. The current QoS profile is fixed by further negotiations between the SGSN and GGSN.
ACTERNA
Page 3-50
GPRS Terminal Equipment Classes GPRS Terminal Equipment Classes

Terminal equipment is split into the following three classes:

Class A terminals Class B terminals

A, B, C
Class C terminals
A, B
GPRS GSM
Page 3-51
A GPRS MS can be operated in one of three different modes. The mode depends on the service for which the MS has attached to the network, and from the ability of the MS to process several services simultaneously or only one at a time. Operating mode class A: The MS is simultaneously GPRS and GSM attached and can utilize both services simultaneously. Operating mode class B: The MS is simultaneously GPRS and GSM attached but can only utilize either the GPRS or the GSM service at a given time. Operating mode class C: The MS is only GPRS attached. The MS cannot handle GSM services.
ACTERNA
Page 3-51
Charging Charging
Determining call charges for GPRS is largely dependent on the operator. Tariffs can be based on

volume or time
or a combination of both.
Page 3-52
The type and method of billing is operator-specific. Every GPRS network operator individually collects and processes the charging information. Most network operators have opted for charging based on volume rather than connection time, while some prefer to use more intelligent methods. They plan to charge according to the type of service used, based on the argument that it is impossible to explain to a customer that it is much more expensive to download a 20 Mbyte e-mail attachment than to book a vacation on line. The cost factor is decisive for the success of GPRS. An extra charging gateway is part of the GPRS portfolio of most infrastructure suppliers, and is intended to give operators the necessary degree of flexibility in determining billing charges. The billing information for each MS is collected in the SGSN and GGSN. Minimum criteria for determining charges are specified in the GPRS specifications for the SGSN and GGSN. Operators can implement their specific billing methods on this basis. SGSN: Recording the data quantity transmitted over the Air interface Time monitoring of an active PDP context Monitoring occupation of general GPRS resources Recording the location (Home PLMN, Visitor PLMN, etc.). GGSN: Recording the telephone numbers involved Recording the data quantity transmitted over external networks Time monitoring of an active PDP context Recording the location (Home PLMN, Visitor PLMN, etc.).
ACTERNA
Page 3-52
Kapitel 4 Kapitel 4
The best organization is often none at all

(Danish proverb)
Page 4-1
Notes:
ACTERNA
Page 4-1
Um Interface Data Um Interface Data

Frequency: Channel spacing: Modulation: Transmission method: Transmission rate per timeslot: Number of timeslots per frame Frame period: Timeslot period: Bit period: 890 to 925 MHz uplink 935 to 960 MHz downlink 200 kHz 0.3 GMSK TDMA T= 4.615 ms 270.833 kbit/s 8 4.615 ms 576.9 us 3.692 us
Page 4-2
The Air interface used for GPRS corresponds to the GSM standard. Transmission method: TDMA (Time Division Multiple Access) together with FDMA (Frequency Division Multiple Access). Every TDMA frame is split into 8 timeslots.
Note: The frequencies given above apply to a GSM 900 system.
ACTERNA
Page 4-2
TDMA Frame Hierarchy TDMA Frame Hierarchy
48
49
50
51
Multiframe
TDMA frame
Burst
Burst
Page 4-3
A Packet Data Channel (PDCH) consists of 52 TDMA frames that are assembled into a multiframe structure. There are 8 timeslots per timeslot that can be used for for burst type data traffic. Timeslot period TDMA frame period Multiframe period = = = 576.9 s 4.615 ms 240 ms
ACTERNA
Page 4-3
Carrier frequency
Physical Channels Physical Channels
fn fn-1 fn-2
TDMA multiframe @ fn TDMA multiframe @ fn-1 TDMA multiframe @ fn-2
...
f2 f1 TDMA multiframe @ f2 TDMA multiframe @ f1
240 ms
Time
Page 4-4
The physical channel is described by the TDMA frame. A single TDMA frame consists of 8 timeslots. 52 consecutive TDMA frames per carrier frequency are collected together into a multiframe. The timeslots within a multiframe can be assigned dynamically or on a fixed basis by the BSS to correspond with the bandwidth requirements of packet data transfer. The frequency hopping procedure familiar from the GSM standard can be used similarly for GPRS. Physical channels in GPRS are generally referred to as Packet Data Channels.
ACTERNA
Page 4-4
Logical Channels (1) Logical Channels (1)

Packet Packet common control common control channels channels (PCCCH) (PCCCH) Packet Packet broadcast control broadcast control channels channels (PBCCH) (PBCCH) Packet Packet dedicated control dedicated control channels channels (PDCCH) (PDCCH) Packet Packet data traffic data traffic channels channels (PDTCH) (PDTCH)
PRACH / /CPRACH PRACH CPRACH

PBCCH / /CPBCCH PBCCH CPBCCH

PACCH PACCH

PDTCH/U PDTCH/U

PPCH / /CPPCH PPCH CPPCH
CFCCH CFCCH
PTCCH/U PTCCH/U
PDTCH/D PDTCH/D
PAGCH / /CPAGCH PAGCH CPAGCH
CSCH CSCH
PTCCH/D PTCCH/D
PNCH / /CPNCH PNCH CPNCH
= Uplink
= Downlink
Page 4-5
The logical GPRS channels at the radio interface can be divided into 4 groups: 1. Packet Common Control Channels 2. Packet Broadcast Control Channels 3. Packet Dedicated Control Channels 4. Packet Data Traffic Channels (C)PRACH (C)PPCH (C)PAGCH (C)PNCH (C)PBCCH CFCCH CSCH PACCH PTCCH PDTCH = = = = = = = = = = (Compact) Packet Random Access Channel (Compact) Packet Paging Channel (Compact) Packet Access Grant Channel (Compact) Packet Notification Channel (Compact) Packet Broadcast Control Channel Compact Frequency Correction Channel Compact Synchronization Channel Packet Associated Control Packet Timing Advance Control Channel Packet Data Traffic Channel
The compact configuration allows operation of cells in which GPRS only and no GSM service is offered. Extra synchronization channels (CFCCH and CSCH) must then be provided for this case. Mapping of the logical channels onto the physical channels is then somewhat different in a compact cell when compared to a normal cell where both circuitswitched and packet-switched services are available.
ACTERNA
Page 4-5
PRACH
PPCH PAGCH PNCH
Packet common control channels
BTS
Page 4-6
The Packet Common Control Channels are used to transmit general signaling messages between the network and the mobile terminal. The access bursts of all MSs are transmitted on the PRACH. Paging messages are transmitted on the PPCH. To enable DRX, paging groups are formed. Paging can be transmitted via this channel for the circuit-switched service (GSM) as well as for the packet-switched service (GPRS). The MS receives notification of the assigned resources (traffic channels, etc.) on the PAGCH before actual data transfer begins. A group of MSs is informed of an impending point-to-multipoint data transfer via the PNCH. The PTM-M service is reserved for future versions and is therefore not specified yet for GPRS Phase 1.
ACTERNA
Page 4-6
PBCCH
CFCCH
Packet broadcast control channels
CSCH
BTS
Page 4-7
The PBCCH is used for transmitting system-specific information (SYS-INFO elements) to all the MSs in a radio cell. If a PBCCH has not been set up, the GPRS-specific system information is transmitted over the usual GSM Broadcast Control Channel BCCH. The CFCCH corresponds to the GSM FCCH. A defined pattern that results in a characteristic line in the frequency spectrum is transmitted over this channel. The mobile station can then synchronize its frequency to this line. Time-synchronization of an MS to the network is accomplished by evaluating the reduced TDMA frame number transmitted on the CSCH. The CSCH is also used to transmit the Base Transceiver Station Identity Code (BSIC). Just like the CFCCH, the CSCH is found in cells that only provide GPRS and where no FCCH and SCH are present as a result.
ACTERNA
Page 4-7
PACCH
PTCCH/U
Packet dedicated control channel
PTCCH/D
BTS
Page 4-8
Specific signaling information for a particular MS is transmitted on the PACCH. The PACCH shares radio resources with the traffic channels. Thus, only mobile stations that have been assigned one or more traffic channels will also have a Packet Associated Control Channel. Paging messages due to incoming circuitswitched GSM connections can then be broadcast to the appropriate mobile over an existing PACCH. The MS continually transmits Access bursts on the PTCCH/U. From these, the BTS determines the current timing advance and communicates this result to the mobile on the PTCCH/D. The MS then advances its transmit time to suit this.
ACTERNA
Page 4-8
PDTCH/U
Packet data traffic channel

PDTCH/D
BTS
Page 4-9
A mobile station is assigned one or more Packet Data Traffic Channels (PDTCH) for transmitting packet data. The user data for the application is then transmitted over these channels. The bandwidth requirement is usually non-symmetrical (an Internet application needs more bandwidth on the downlink than on the uplink to download graphics or programs, for example). As a result, the MS is assigned different resources for uplink and downlink (in contrast, the uplink and downlink resource assignments are strictly symmetrical for the circuit-switched GSM service). An MS can be assigned several PDTCHs for a data transfer and can use these in parallel to increase the bandwidth. This is called multislot operation, and must of course also be supported by the terminal equipment. All PDTCHs are unidirectional. The PDTCH/U is used for uplink transfers and the PDTCH/D for downlink transfers. This is logical considering the non-symmetrical bandwidth requirements.
ACTERNA
Page 4-9
Multiframe Structure of a PDCH Multiframe Structure of a PDCH

52 TDMA frames
B0
B1
B2
B3
B4
B5
B6
B7
B8
B9
B10
B11
4 TDMA frames / radio block 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7
B0 - B11 = Radio blocks T = Frame for timing advance (PTCCH) X = Idle frame
Page 4-10
A PDCH multiframe is split into 12 radio blocks. Each block in itself consists of 4 consecutive TDMA frames. Two T frames and two idle frames are placed between each block. The Timing Advance Control Channels (PTCCH) are mapped on to the T frames. The frequency correction and synchronization channel (CFCCH and CSCH) are transmitted in the idle frames in the compact configuration. When the physical resources are assigned to the logical channels, radio blocks 0 through 11 are occupied in the following prescribed order: B0 - B6 - B3 - B9 - B1 - B7 - B4 - B10 - B2 - B8 - B5 - B11 The exact mapping of the logical GPRS channels in to the PDCH multiframes is described by various parameters, which are transmitted continually on the PBCCH.
ACTERNA
Page 4-10
Downlink Mapping Example Downlink Mapping Example

BCCH BCCH (GSM) (GSM)
B0 B1 B2 T B3 B4 B5 X B6 B7 B8 T B9 B10 B11 X
PBCCH (optional)
PDCH a
PBCCH Further logical downlink channels PTCCH/D
PBCCH (optional)
PDCH b
B0 B1 B2 T B3 B4 B5 X B6 B7 B8 T B9 B10 B11 X Further logical downlink channels PTCCH/D Further logical downlink channels Further logical downlink channels PTCCH/D Further logical downlink channels
Further logical downlink channels
Further logical downlink channels PTCCH/D

PBCCH (optional)
Page 4-11
The GSM BCCH points to a PDCH, which contains the PBCCH. This is mandatorily mapped in to block 0 and can occupy up to 3 further blocks (6, 3, 9) if necessary. The PDCH containing the Packet Common Control Channels (PCCCH) is indicated in the PBCCH. The PCCCHs can lie on the same PDCH as the PBCCH. A maximum of 12 blocks (in the prescribed order of B0-B6-B3-) in all PDCHs are filled with the logical channels PAGCH, PNCH, PDTCH, and PACCH. The blocks remaining after this can be filled with the channels PPCH, PAGCH, PNCH, PDTCH and PACCH. In any case, the transmitted message type determines the actual use of a radio block. In all, this mapping procedure is very dynamic. Many parameters that are continually broadcast in the PBCCH determine the actual mapping. The diagram above should thus be seen as representing one of many different possibilities. The assignment and occupation of PDTCHs and PACCHs takes place within the framework of a packet data transfer between the network and the MS. A PDTCH is precisely mappedon to a physical channel (frequency, timeslot). Up to 8 PDTCHs with different timeslots but the same frequency can be assigned to an MS at the same time. The PACCH lies on the same physical channel as the PDTCH in this case. The current usage of the physical channel (as PDTCH or as PACCH) is indicated by the message transmitted over the PDCH. Note: The PBCCH is not always required. In such cases, the corresponding parameters must be broadcast in the BCCH directly.
ACTERNA
Page 4-11
Uplink Mapping Example Uplink Mapping Example

BCCH BCCH (GSM) (GSM) Downlink PDCH
B0 B1 B2 T B3 B4 B5 X B6 B7 B8 T B9 B10 B11 X
PBCCH (optional)
Uplink PDCH
B0 B1 B2 T B3 B4 B5 X B6 B7 B8 T B9 B10 B11 X PRACH PDTCH/U or PACCH PTCCH/U PRACH PDTCH/U or PACCH PRACH PDTCH/U or PACCH PTCCH/U PRACH PDTCH/U or PACCH
PBCCH Further logical downlink channels PTCCH/D

PBCCH (optional)

PBCCH (optional)
Page 4-12
In the uplink direction, all blocks may be assigned the logical channels PRACH, PDTCH/U and PACCH. If the uplink PDCH also contains the PRACH, the first n blocks will be exclusively occupied by this according to the B0-B6-B3-... sequence. The mapping parameters of the logical signaling channels are broadcast on the PBCCH in the downlink.
ACTERNA
Page 4-12
Timeslot Mapping for Compact Cells Timeslot Mapping for Compact Cells
TDMA frame TS 0 TS 1 TS 2 TS 3 TS 4 TS 5 TS 6 TS 7
CPBCCH CPBCCH CPRACH CPRACH CPPCH CPPCH CPAGCH CPAGCH CPNCH CPNCH
PDTCH PDTCH PACCH PACCH
Page 4-13
For compact cells, the synchronization information (CFCCH and CSCH) and the CPBCCH are transmitted on a special carrier frequency. This specific frequency within the mobile allocation is comparable with the BCCH frequency of a normal GSM cell. Compact cells have a mobile allocation of typically (at least) 3 carrier frequencies. The frequency use do transmit the synchronization information and the CPBCCH is called the Primary Compact Carrier, all others are referred to as Secondary Compact Carrier(s). The CPBCCHs are only transmitted on the primary carrier. All other channels are transmitted on the primary and / or secondary carriers.
ACTERNA
Page 4-13
Compact Downlink Mapping Example Compact Downlink Mapping Example

Primary compact carrier
B0 B1 B2 T B3 B4 B5 X B6 B7 B8 T B9 B10 B11 X CPBCCH Further logical downlink channels PTCCH/D
CPBCCH (optional)
Secondary compact carrier(s)

B0 B1 B2 T B3 B4 B5 X B6 B7 B8 T B9 B10 B11 X Further logical downlink channels PTCCH/D Further logical downlink channels Further logical downlink channels PTCCH/D Further logical downlink channels
Further logical downlink channels CFCCH bzw. CSCH

CPBCCH (optional)

CPBCCH (optional)
Further logical downlink channels CFCCH bzw. CSCH
Page 4-14
Frequency and time synchronization information is transmitted in the idle frames on the Primary Compact Carrier. The logical compact channels may be distributed in the following order over the radio blocks: On the Primary Carrier only: CPBCCH on block 0 and optionally on blocks 6, 3 and 9 Subsequently, the up to 12 blocks (in the prescribed order of B0-B6-B3-B9-) are occupied with the channels CPAGCH, CPNCH, PDTCH and PACCH on all carriers (primary und secondary). Any blocks now remaining can be filled with the channels CPPCH, CPAGCH and CPNCH. In any case, the transmitted message type determines the actual use of a radio block. The current distribution of the logical channels on the radio block is described by a set of parameters that is continually broadcast on the CPBCCH.
ACTERNA
Page 4-14
Compact Uplink Mapping Example Compact Uplink Mapping Example

Primary or secondary compact carrier
B0 B1 B2 T B3 B4 B5 X B6 B7 B8 T B9 B10 B11 X CPRACH Further logical uplink channels PTCCH/U CPRACH Further logical uplink channels CPRACH Further logical uplink channels PTCCH/U CPRACH Further logical uplink channels
Page 4-15
Compact uplink mapping corresponds for the most part to the uplink mapping already described. However, the CPRACH is used here instead of the PRACH. This can be present on the primary and on all the secondary carriers.
ACTERNA
Page 4-15
Principle of Radio Resource Management Principle of Radio Resource Management

Distribution of the available physical channels of a cell between GSM and GPRS services is a dynamic process. Master - Slave concept Capacity on demand concept - Perform load monitoring - Use idle channels as PDCHs - Take higher priority network services into account
Rapid freeing up of occupied GPRS channels
Page 4-16
A flexible, dynamic utilization of resources that is oriented towards the momentary requirements should be possible for the circuit-switched (GSM) and packet-switched (GPRS) services existing in a cell. Master Slave concept: The signaling required to set up a packet connection (Packet Common Control Channels and Packet Broadcast Control Channels) is transmitted on a PDCH (master). Then only the PTCCHs and PACCHs are transmitted on all other PDCHs (slaves). Capacity on demand concept: The network operator determines the proportion of radio resources to be occupied by GPRS in accordance with the GPRS traffic load (load monitoring). It is not necessary that a PDCH is occupied permanently. If no physical channel is reserved for GPRS, the MS must listen on the BCCCH. The MS then indicates a Packet Request on the RACH instead of the PRACH. The network then reserves a GPRS PDCH for the imminent packet data transfer and indicates this to the mobile. Radio resources that are not currently in use can be assigned to GPRS to improve the QoS (bandwidth, transmission quality, etc.) of the service. If higher priority services (e.g. GSM circuit-switched services) require resources, the GPRS PDCHs must be capable of being released rapidly. Release of a PDCH is indicated to the affected MS by the network in the form of a broadcast message.
ACTERNA
Page 4-16
Frequency
Resource Distribution Example Resource Distribution Example

5 6 7 0 1 2 3 4 5 6 7 0 1 2
Timeslot no.
Frequenz 3 f1 f2 f3 f4 4
Downlink
3 f1 f2 f3 f4
Uplink
Timeslot-Nr.
GSM-Teiln. 1 GSM-Teiln. 2
GPRS-Teiln. 1 GPRS-Teiln. 2 GPRS-Teiln. 3

Page 4-17
The radio resources can be split between GSM circuit-switched and GPRS packet-switched subscribers. GSM subscribers use circuit-switched services and thus occupy fixed timeslots. The timeslot numbers are always identical for uplink and downlink, i.e. resources are occupied symmetrically. GPRS subscribers can be assigned several timeslots within a radio block from the network. The uplink resource assignments will normally differ from the downlink assignments. Usually, more resources are needed for the downlink than for the uplink.
ACTERNA
Page 4-17
Layer Model at the Um Interface Layer Model at the Um Interface

Higher protocol layers Higher protocol layers
RLC
L2
RLC MAC
MAC
Physical link
L1
Physical link Physical RF

Network Um
Physical RF
MS
Page 4-18
The physical layer at Um is split into two part layers, the Physical Link Layer and Physical RF Layer (Radio Frequency). Tasks and functions of the Physical RF Layer: Providing the physical channels corresponding to GSM 05.02 (FDMA/TDMA multiple access) Modulating the data to be transmitted (GMSK modulation) Tasks and functions of the Physical Link Layer: Forward Error Correction (FEC) by means of appropriate channel coding methods Splitting a radio block across 4 bursts and transmitting these bursts in consecutive TDMA timeslots Perform interleaving Determining overload states Timing Advance functions Monitoring transmission quality Cell re-selection functions (comparable to GSM handover) Power control functions Power save functions to save battery power for the MS (DRX)
ACTERNA GPRS Radio Interface Page 4-18
Radio Block Structure Radio Block Structure

Radio block for data transfer
MAC header RLC header RLC data BCS
Channel coding
Normal burst Normal burst Normal burst Normal burst
Channel coding
MAC header RLC/MAC control message BCS
Radio block for signaling
Page 4-19
Two different block structures are defined for data and RLC / MAC signaling messages. The MAC header contains different control fields for uplink and downlink. The length is 8 bits. The RLC header contains different control fields for uplink and downlink. The length is variable. The Block Check Sequence (BCS) is used to detect errors on the transmission path. The data of the higher protocol layers (in this case the RLC layer) are transmitted in the RLC data field. The RLC / MAC Control Message field contains a signaling message from the RLC / MAC layer. 114 user bits can be transmitted per normal burst. After channel coding, the length of a radio block is therefore 456 bits.
ACTERNA
Page 4-19
Channel Coding (1) Channel Coding (1)

Radio block
USF Convolutional coding
BCS
Punctuation
456 bits
Coding schemes CS-1, CS-2, CS-3
Page 4-20
Coding schemes CS-1, CS-2 and CS-3 include sufficient redundancy to be used by the receiver for a forward error correction procedure, thanks to the convolution code that is used. The radio block is first subjected to convolution coding. The uplink state flag (USF) is part of the MAC header and is coded separately from the rest of the data. With CS-2 and CS-3, the result of complete coding is greater than the permitted maximum of 456 bits. For this reason, unimportant bits are removed again from the coded data block that is formed. The coded USF is not included in this. The CS-4 coding scheme does not generate any code redundancy and so no forward error correction (FEC) is possible. CS-1 is used on all signaling channels except PRACH and PTCCH/U.
ACTERNA
Page 4-20
Channel Coding (2) Channel Coding (2)
Radio block
Block coding
USF
BCS
No coding
456 bits
Coding scheme CS-4
Page 4-21
GPRS coding scheme parameters:
ACTERNA
Page 4-21
Cell Reselection Cell Reselection
Mobile controlled Network controlled

or
Page 4-22
In GPRS operation, cell reselection is normally carried out by the MS. The MS monitors the receive field strengths of its own and the neighbor cells on the BCCH carrier or on the CPBCCH carrier. The neighbor cells are recognized by comparing the BSICs received there with the values from the neighbor cell list which is broadcast over the current broadcast channel. Once an MS has selected a new cell which can be used for GPRS and for which appropriate resources have been assigned, the packet data transfer ends in the old cell and is resumed in the new cell. As well as this normal cell reselection procedure, the MS can also be requested to relinquish this facility temporarily and accept a cell reselect decision from the network. The network may request explicit measurement results from the MS for this purpose. If a class A mobile station is running GPRS and GSM connections simultaneously, the GSM handover has priority over the GPRS cell reselection procedure, regardless of whether this is controlled from the mobile or from the GPRS network.
ACTERNA
Page 4-22
Timing Advance Timing Advance
Initial determination of TA
Continuous updates of TA
PTCCH/U Packet channel request
PTCCH/D Packet uplink / downlink assignment
BTS BTS
Page 4-23
Initial determination of TA: The MS sends a Packet Channel request in the form of an access burst on the PRACH. The network calculates the current TA from the delay and indicates this to the mobile together with the resource assignment by means of the Packet Uplink Assignment or Packet Downlink Assignment messages. Continuous TA updates: The TA information is transmitted over assigned sub-channels of the Packet Timing Advance Control Channel (PTCCH). The Timing Advance Index is transmitted as an additional parameter in the Packet Uplink / Downlink Assignment message. This index indicates the sub-channel of the PTCCH that the mobile station is to use. The MS continually transmits access bursts in the assigned PTCCH on the uplink, which the network can use to determine the current timing advance. The network indicates the current TA value to the MS on the downlink PTCCH in the form of a signaling message. This is transmitted by means of normal bursts.
ACTERNA
Page 4-23
Wait for MS data
Power Control Power Control

Basic SDL flowchart for the BTS
MS data
Downlink measurement values? no Perform uplink measurements
yes
Adjust BTS output level
w Do
k in nl
ea m
m re su
t en
s ue al v
MS power command
po
om rc e
an m
BTS
Wait for MS data
Page 4-24
Power control is used to achieve better spectral characteristics (fewer interferers, etc.) and economical use of the battery capacity of the MS. The BTS indicates to the MS the value of the output level that it must set. The BTS determines this from the results of the uplink measurement. The BTS calculates its output level from the downlink measurement results that are obtained by the MS and then transmitted to the BTS. The BTS uses a constant level on all PDCHs with PBCCH and PPCH. This level can be less than the level at the BCCH frequency. The level difference is indicated in the PBCCH. The MS performs its level measurement on the PCCCH or on the BCCH if the PCCCH is not present.
ACTERNA
Page 4-24
Activities During TA and IDLE Frames Activities During TA and IDLE Frames
The MS performs the following tasks during IDLE frames or PTCCH frames:

BSIC detection for cell reselection
Continuous timing advance procedure
Interference measurements for power control
Page 4-25
Timing Advance information is transmitted on the PTCCH. In case of compact configuration the BSIC detection runs during the idle frames. Interference measurement can be performed from a MS in packet transfer mode on idle frames as well as on the PTCCH frames. The channels on which a MS in packet idle mode is to perform interference measurements are indicated to the MS in the PBCCH, or, if this is non-existent, in the BCCH. If no channels are specified there, the MS may not perform any interference measurements.
ACTERNA
Page 4-25
Chapter 5 Chapter 5
RLC / MAC Radio Link Control / Medium Access Control
The top leads while the bottom bleeds

(Martin Kessel)
Page 5-1
Notes:
ACTERNA
MAC / RLC
Page 5-1
Layer Model Layer Model

Application IP/X.25 SNDCP (or MM) LLC RLC RLC MAC GSM RF MAC GSM RF Relay BSSGP BSSGP Network service L1bis Network service L1bis IP L2 L1 IP L2 L1 Relay GTP SNDCP/MM LLC GTP UDP/ TCP UDP/ TCP IP/X.25
MS
Um
BSS
Gb
SGSN
Gn
GGSN
Gi
Page 5-2
The MAC function provides the procedures necessary for several mobile terminals to be able to use a common transmission medium. The latter can consist of several physical channels. The MAC layer also provides paging services for mobile terminals for the higher protocol layers. The RLC / MAC layer can transport user data in acknowledged and unacknowledged modes. In contrast with unacknowledged mode, the acknowledged mode allows repetition of damaged data blocks in the RLC / MAC layer, thus providing protected operation. MAC also allows parallel use of several physical channels by one mobile station (multislot mode).
ACTERNA
MAC / RLC
Page 5-2
Protocol Architecture Protocol Architecture

Mobility management (MM) sub-layer Mobility management (MM) sub-layer LLC sub-layer LLC sub-layer
RR-SAP GMMRR-SAP GRR-SAP
Radio resource (RR) sub-layer
Radio resource (RR) management

RR Higher layer PDUs
RLC / MAC
PBCCH PCCCH PACCH PDTCH
SDCCH
SDCCH
SACCH
SAPI 0
SACCH
FACCH
PCH AGCH
BCCH
RACH
SAPI 3
Data link layer Data link layer PDCH Physical link layer Physical link layer
Page 5-3
The RR layer covers the following functions: Management of radio resources (RR Management). Management of logical connections (RLC). Making available the necessary functions for transmitting Protocol Data Units (PDUs) of higher layers (RLC). Backward error correction (RLC) Monitoring the correct order of PDUs (RLC) Segmenting and reassembly of the LLC PDUs to be transmitted (RLC). Performing paging, multiple attempts, etc. (MAC) Organize and manage access of different MSs to common physical resources (MAC) Collision detection and clearance during simultaneous access of the common radio medium by several mobile stations (MAC).
RLC = Radio Link Control MAC = Medium Access Control
ACTERNA
MAC / RLC
Page 5-3
GPRS Radio Resource Management A mobile terminal that is making use of GPRS is assigned a number of physical channels for packet communications. The GPRS signaling initially takes place in the PCCCH (Packet Common Control Channel) or in the CCCH (Common Control Channel). At least one Packet Data Channel (PDCH) functions here as master in order to accommodate the necessary signaling in the corresponding PCCHs. The assignment of physical channels for GPRS and GSM takes place according to the principle of capacity on demand. This means that GPRS PDCHs are allocated as required. The signaling for making PDCHs available takes place over the PCCCH if it exists. Otherwise, a CCCH takes over this task. The availability of a packetspecific signaling channel (PCCCH) is indicated to the existing terminals via the BCCH.
Capacity on demand means that the number of PDH channels can change according to need. The following measures are taken to allow this: Monitoring of the load status of the network (in the cell). This task is performed by MAC (Medium Access Control). Dynamic allocation of PDCHs. Traffic channels that are not in use can be used as PDCHs.
Radio Resource Management recognizes the following two operating modes: 1. Packet idle mode 2. Packet transfer mode
ACTERNA
MAC / RLC
Page 5-4
RLC //MAC Frame Structure for User Data RLC MAC Frame Structure for User Data
7 6 5 RRBP 4 3 S/P 2 1 USF 0 Downlink
Payload type
Payload type
Countdown value
SI
Uplink
MAC header
RLC header
RLC data (LLC-PDUs)
Spare
BCS
PR BSN Spare BSN 7 6 5 4
TFI
FBI E
Downlink Byte 1 Downlink Byte 2 Uplink Byte 1 Uplink Byte 2
TFI
TI E
Page 5-5
Payload Type: Describes the type of data contained in the block (user data or signaling). Payload Type 0 = user data. RRBP: Relative Reserved Block Period. With this, the network indicates to the MS the radio block (B0 through B11) in which it can transmit PACKET CONTROL ACKNOWLEDGEMENT or PACCH messages. The same timeslot should be used in the uplink as the RRBP information was received on in the downlink. S/P: Supplementary/Polling Bit (0: RRBP is not valid, 1: RRBP is valid) USF: Uplink State Flag. The USF is continually transmitted in the downlink and indicates which MS may next transmit its queued uplink block (traffic light function). Countdown Value: Indicated with every uplink transmission of a data block. This enables the network to estimate the remaining duration of the packet data transfer (or so-called temporary block flow). The Countdown Value is determined using a special calculation rule. SI: Stall Indicator. Indicates whether the time of MS the transmit window is fixed or continuous. The MS should set the SI bit to 1 (fixed time window) in all uplink data blocks. R: Retry Bit. Used in conjunction with a Channel Request, indicating to the network whether this was the first or a subsequent Channel Request attempt by the MS.
ACTERNA
MAC / RLC
Page 5-5
PR: Power Reduction. Refers to the reduced level value of the current RLC blocks compared to the BCCH level. TFI: Temporary Flow Identifier. Indicates that a RLC block belongs to a certain temporary block flow. FBI: Final Block Indicator. Indicates the last block ina downlink TBF (FBI = 1 for the last block). BSN: Block Sequence Number. Incremented each time a RLC / MAC block is transmitted. The BSN starts from 0 and counts in modulo 128. E: Extension Bit. E = 0 means that a further (optional) byte follows. Spare: Unused bits. TI: TLLI Indicator. Indicates whether an optional TLLI field is present in the RLC data block or not. The TLLI (Temporary Logical Link Identifier) is used by the LLC protocol. BCS: Block Check Sequence. RLC data: The RLC data field contains octets from one or more LLC PDUs.
Size of RLC data blocks according to coding scheme:

but including RLC header
ACTERNA
MAC / RLC
Page 5-6
RLC Modes RLC Modes

Acknowledged / Unacknowledged Mode
BTS
User data block
User data block (last in TX window)
PACKET DOWNLINK ACK/NACK
Page 5-7
The RLC layer provides the possibility of protected transport of user data blocks. In this operating mode, the block counter BSN is incremented each time the transmit side transfers a new block. The receive side uses the BSN to acknowledge correct receipt of the blocks and to ensure that the order is correct. The acknowledgement must be returned from the receive side to the transmit side within the prescribed transmit window by means of the PACKET DOWNLINK ACK/NACK or PACKET UPLINK ACK/NACK messages. The standard window size is 64 data blocks. Acknowledged operating mode allows detection of blocks that are incorrectly transmitted by the RLC layer, and their correction by repeating the frame, as well as the repeating of lost blocks.
Unacknowledged mode does not include any provisions for repeating user data blocks. Error correction in the RLC layer is thus not provided. The block number BSN in the RLC header is only needed to ensure that the data blocks are received in the correct order.
ACTERNA
MAC / RLC
Page 5-7
RLC //MAC Frame Structure for Signaling RLC MAC Frame Structure for Signaling
7 6 5 RRBP 4 3 S/P 2 1 USF 0 Downlink
Payload type
Payload type
Spare
Uplink
MAC header
RLC header
RLC / MAC signaling
BCS
RBSN Spare 7 6 5
RTI TFI 4 3 2
FS
AC D
Downlink Byte 1 (optional) Downlink Byte 2 (optional)
Page 5-8
The RLC header is only present in the downlink and is wholly or partially optional there. The presence or absence of the RLC header can be determined from the payload type coding. RBSN: Reduced Block Sequence Number. Count in modulo 2. RTI: Radio Transaction Identifier. Serves to logically distinguish or separate several RLC connections operating in parallel. FS: Final Segment. Indicates the last segment of a downlink RLC / MAC signaling message (1 = last segment). AC: Address Control bit. Indicates whether the subsequent optional TFI/D byte is present (1 = present). D: Direction bit. Indicates the direction of the temporary block flow referred to by the TFI (0 = uplink TBF, 1 = downlink TBF).
ACTERNA
MAC / RLC
Page 5-8
Multiplex Principle Multiplex Principle

Multiplex formation in the RLC/MAC layer is based on the following procedure:

Temporary block flow Temporary flow identity Uplink state flag Medium access modes
Page 5-9
Temporary Block Flow (TBF) A physical connection of two RR instances (between MS and network) performing an exchange of LLC PDUs is called a Temporary Block Flow. Such a TBF is temporarily assigned to an exchange of LLC packets. Temporary Flow Identity (TFI) Every TBF has a TFI identifying it. A TFI value has the same meaning as an MS identity in the RLC / MAC layer, i.e. every TFI only occurs once in one direction (uplink or downlink) at a given point in time. Uplink State Flag (USF) The USF is used by the network to (dynamically) control access of different mobile stations to the transmission medium. Medium access modes The following three access modes are defined for GPRS networks: Dynamic allocation Extended dynamic allocation Fixed allocation All networks must support at least dynamic allocation and fixed allocation. Extended dynamic allocation is optional.
ACTERNA
MAC / RLC
Page 5-9
The three different ways that the network can assign uplink resources to the mobile station are described below:
1. Dynamic allocation The network assigns uplink transmission resources to the mobile station in a completely dynamic manner (on-line). For this, the MS is first assigned a USF for each PDCH in addition to the PDCHs. The MS now continually monitors the USF in the corresponding downlink PDCH for every allocated uplink PDCH and compares this value with the assigned USF. If the two are the same, the MS can then transmit its data over the corresponding uplink PDCH. To enable the mobile to monitor the USF in the downlink, RLC data blocks must be transmitted continually in the downlink channels. If the network does not have any data or signaling blocks to transmit, the RLC / MAC signaling message PACKET DOWNLINK DUMMY CONTROL BLOCK can be transmitted as a filler message.
2. Extended dynamic allocation Basically the same as dynamic allocation. The allocation is extended in the sense that the mobile station can transfer data on all further assigned PDCHs with higher sequence numbers instead of on only one PDCH after a positive USF comparison. If a mobile supports extended dynamic allocation, it indicates this to the network in the PACKET RESOURCE REQUEST: MS Radio Access Capability message.
3. Fixed allocation The corresponding PACKET UPLINK ASSIGNMENT message only contains fixed allocations of radio blocks (B0 through B11) and PDCHs (frequencies, timeslots) for uplink data transfer. The USF has no meaning here. At the same time, the network indicates the PACCH reserved in the downlink in the PACKET UPLINK ASSIGNMENT message to the MS.
ACTERNA
MAC / RLC
Page 5-10
Packet Idle Mode and Packet Transfer Mode Packet Idle Mode and Packet Transfer Mode
LLC-PDU transfer request
Packet idle mode Packet idle mode

-- no resources occupied no resources occupied -- no TBF exists no TBF exists
Packet transfer mode Packet transfer mode

-- Resources occupied Resources occupied -- at least one TBF exists at least one TBF exists
LLC-PDU transfer ready
Page 5-11
There is no flow of data packets (Temporary Block Flow = TBF) in packet idle mode. The higher layers can initiate TBF by transmitting an LLC PDU. This causes the transition to packet transfer mode. The mobile terminal monitors the PBCCH (Packet Broadcast Control Channel) and the paging channel in packet idle mode. If there is no PBCCH, the BCCH is monitored instead. In this state, a class A terminal can utilize other GSM services. Class B and class C terminals must first leave the GPRS states (idle mode or transfer mode). In packet transfer mode, the mobile terminal has radio resources assigned to it, i.e. a temporary block flow (TBF) is possible in one or more physical channels. Several TBFs can be set up simultaneously in different directions. The RLC layer allows acknowledged or unacknowledged transfer of LLC PDUs. Before transferring to a new cell, the MS goes into packet idle mode, switches to the new cell and then reverts to packet transfer mode. A class A terminal can also utilize other GSM services in this mode.
ACTERNA
MAC / RLC
Page 5-11
Mobile Originated Packet Transfer Mobile Originated Packet Transfer
BTS
PACKET CHANNEL REQUEST PRACH (or RACH)
PACKET UPLINK ASSIGNMENT PAGCH (or AGCH)
PACKET RESOURCE REQUEST PACCH optional PACKET UPLINK ASSIGNMENT PACCH
Page 5-12
The terminal starts the procedure with a PACKET CHANNEL REQUEST message in the (Packet) Random Access Channel (PRACH or RACH). This message includes the required resources. This request is acknowledged with a PACKET UPLINK ASSIGNMENT message which confirms reservation of the required resources. Up to 8 RLC data blocks can be registered for an uplink transfer with the PACKET CHANNEL REQUEST message. If this capacity is insufficient, a so-called twophase access request is made in the PACKET CHANNEL REQUEST message as before. In this case, further information about the required resources is now transmitted in the Packet Associate Control Channel (PACCH) in the form of a PACKET RESOURCE REQUEST message. This is then acknowledged by the second PACKET UPLINK ASSIGNMENT message with the release of further resources. If an MS needs no more than 8 RLC data blocks, the PACKET CHANNEL REQUEST message carries the label Short Access Request. The Packet Uplink Assignment message contains A list of the data channels (PDCH) The Uplink State Flag (USF) pertaining to each PDCH to enable multiplexing of the RLC blocks from different terminals. TFI value
ACTERNA
MAC / RLC
Page 5-12
Packet Uplink Assignment Message (1) Packet Uplink Assignment Message (1)
< Packet Uplink Assignment message content > ::= < PAGE_MODE :bit (2) > {0 | 1 <PERSISTENCE_LEVEL :bit (4) >} { 0 < Global TFI : < Global TFI IE > > |10 < TLLI : bit (32) > | 110 < TQI : bit (16) > | 111 < Packet Request Reference : < Packet Request Reference IE > > } < CHANNEL_CODING_COMMAND :bit (2) > < TLLI_BLOCK_CHANNEL_CODING :bit (1) > < Packet Timing Advance : < Packet Timing Advance IE > > {0 | 1 < Frequency Parameters : < Frequency Parameters IE > > } {01 <Dynamic Allocation : < Dynamic Allocation structure > > |10 <Single Block Allocation : < Single Block Allocation structure > > | 00 < extension > |11 < Fixed allocation : < Fixed Allocation structure > > }
Page 5-13
An extract from a PACKET UPLINK ASSIGNMENT message is shown above and on the two following pages. If a parameter is preceded by 0|1, it is optional. The most important parameters in this message are: Global TFI: If this message refers to an existing TBF, the TBF index (unique TBF number) is shown here. An uplink or downlink TFI can be specified here. Packet Request Reference: Contains the reference for the preceding PACKET CHANNEL REQUEST. CHANNEL_CODING_COMMAND: Indicates the coding scheme (CS-1, CS-2, etc.). Packet Timing Advance: Contains the value, index (TAI) and timeslot for uplink and downlink separately. Frequency Parameters: Contains information about the assigned frequencies (ARFCN, frequency hopping, etc.). Dynamic Allocation: Information about dynamic resource allocation. Fixed Allocation: Information about fixed resource allocation.
ACTERNA
MAC / RLC
Page 5-13
<Dynamic Allocation structure > ::= < Extended Dynamic Allocation :bit (1) > {0 | 1 < P0 :bit (4) > } < USF_GRANULARITY :bit (1) > {0 | 1 < UPLINK_TFI_ASSIGNMENT :bit (5) > } {0 | 1 < RLC_DATA_BLOCKS_GRANTED :bit (8) > } {0 | 1 < TBF Starting Time : < Starting Frame Number Description IE > > } {0 -- Timeslot Allocation {0 | 1 < USF_TN0 :bit (3) > } {0 | 1 < USF_TN1 :bit (3) > } {0 | 1 < USF_TN2 :bit (3) > } {0 | 1 < USF_TN3 :bit (3) > } {0 | 1 < USF_TN4 :bit (3) > } {0 | 1 < USF_TN5 :bit (3) > } {0 | 1 < USF_TN6 :bit (3) > } {0 | 1 < USF_TN7 :bit (3) > }}
Page 5-14
Extended Dynamic Allocation : 0 = dynamic allocation, 1 = extended dynamic allocation. P0: Power Control Parameter for the downlink. USF_GRANULARITY: In the case of dynamic allocation, the MS is informed whether it can send only one or 4 consecutive RLC/MAC blocks after it has recognized its USF. UPLINK_TFI_ASSIGNMENT: A unique TFI for the subsequent TBF is assigned here. RLC_DATABLOCKS_GRANTED: Specifies a fixed number of RLC data blocks that may be sent in the subsequent TBF (9 to 264). If this information element is missing, the duration of the TBF is undefined. TBF Starting Time: Contains the absolute TDMA frame number with which the TBF starts. The MS does not need to monitor the downlink USF prior to this frame number. USF_TNx: The USF values are assigned to each timeslot separately here. Note: The leading bit (0 or 1) indicates if this timeslot will be assigned or not.
ACTERNA
MAC / RLC
Page 5-14
<Fixed Allocation structure > ::= {0 | 1 < UPLINK_TFI_ASSIGNMENT :bit (5) > } < FINAL_ALLOCATION :bit (1) > < DOWNLINK_CONTROL_TIMESLOT:bit (3) > {0 | 1 < P0 :bit (4) > < BTS_PWR_CTRL_MODE :bit (1) > } {0 < TIMESLOT_ALLOCATION :bit (8) > |1 < Power Control Parameters : < Power Control Parameters IE > > } < HALF_DUPLEX_MODE :bit (1) > < TBF Starting Time : < Starting Frame Number Description IE > > {0 {0 -- with length of Allocation Bitmap < BLOCKS_OR_BLOCK_PERIODS :bit (1) > < ALLOCATION_BITMAP_LENGTH :bit (7) > < ALLOCATION_BITMAP : bit (val(ALLOCATION_BITMAP_LENGTH)) > |1 -- without length of Allocation Bitmap (fills remainder of the message) < ALLOCATION_BITMAP :bit ** >} ! < Message escape : 1 bit (*) = <no string> > } ;
Page 5-15
UPLINK_TFI_ASSIGNMENT: A unique TFI is assigned for the subsequent TBF here. FINAL_ALLOCATION: 0 = further allocation messages follow, 1 = resource allocation completed. DOWNLINK_CONTROL_TIMESLOT: Indicates to the MS the downlink TS in which the PACCH is transmitted. TIMESLOT_ALLOCATION: Every bit set in this byte corresponds to an allocated timeslot. ALLOCATION_BITMAP: Contains complete information about the radio blocks (B0 through B11) during which the MS may transmit its uplink data in the timeslots assigned to it.
ACTERNA
MAC / RLC
Page 5-15
Example of a Fixed Allocation Example of a Fixed Allocation

Allocated timeslots
B0
B1
B2
B3
B0
B1
B2
B3
... ... ... ... ...
B9
B10
B11
B9
B10
B11
B0
B1
B2
B3
B9
B10
B11
B0
B1
B2
B3
B9
B10
B11
B0
B1
B2
B3
B9
B10
B11
Radio blocks
Page 5-16
In this example, timeslots 0, 1, 2, 5 and 7 are assigned to the MS for uplink transfer. Blocks of time are defined for each assigned timeslot, during which the MS may transfer its data in the timeslots so assigned. The MS is informed of a possible change in resource assignment via the downlink PACCH (which was indicated in the fixed allocation information element).
ACTERNA
MAC / RLC
Page 5-16
Uplink RLC Data Block Transfer Uplink RLC Data Block Transfer
Data block Data block Data block (last in TX window) Packet Uplink ACK/NACK (NACK) Data block (repeat) Data block Data block Packet Uplink Assignment (change of resources) Packet Control Acknowledgement Data block Data block (last in the TBF) Packet Uplink ACK/NACK (Final ACK)
PDTCH PACCH PDTCH PACCH
BTS
PDTCH
Page 5-17
Example of a message sequence during an uplink packet data transfer. This is, of course, preceded by the resource assignment by means of a Packet Uplink Assignment message. After receiving n data blocks (transmit window) the network acknowledges correct (or incorrect) reception. If transmission errors have occurred, the last m data blocks will be repeated. In the example above, a change in resources is made during an active TBF by means of the Packet Uplink Assignment message. The MS acknowledges with a Packet Control Acknowledgement. After the last data block has been received, the network acknowledges this with Packet Uplink Ack/Nack (Final Ack) and then releases the TBF and the other resources. If no further TBF exists, the MS reverts to packet idle mode. The network recognizes the last data block by means of the count down counter in the MAC header, or a fixed number of data blocks was agreed in the RLC_DATABLOCKS_GRANTED information element.
ACTERNA
MAC / RLC
Page 5-17
Mobile Terminated Packet Transfer Mobile Terminated Packet Transfer
BTS PACKET PAGING REQUEST PPCH (or PCH) PACKET CHANNEL REQUEST PRACH (or RACH) PACKET UPLINK ASSIGNMENT PAGCH (or AGCH) PACKET RESOURCE REQUEST PACCH PACKET UPLINK ASSIGNMENT PACCH PACKET PAGING RESPONSE PDTCH
LLC frame
Page 5-18
Initially, the network sends one or more paging messages to the mobile. After this, the necessary uplink resources are assigned. In the example above, the mobile performs a two phase access request. Once the uplink resources have been allocated, the MS can respond to the paging message. Since the PACKET PAGING RESPONSE message is a data frame in the LLC layer, it is transferred in a PDTCH (payload type = data). The RLC/MAC instance on the network side passes the received PACKET PAGING RESPONSE message on to the superior LLC protocol instance.
ACTERNA
MAC / RLC
Page 5-18
Downlink RLC Data Block Transfer Downlink RLC Data Block Transfer
1. 2. 3. 4. 5. 6. 7. 8. 9. Packet Downlink Assignment Data block Data block Data block (polling) Packet Downlink ACK/NACK Data block Data block Packet Downlink Assignment (polling) Packet Control Acknowledgement
BTS
10. Data block 11. Data block (last of TBF, polling) 12. Packet Downlink ACK/NACK (Final ACK)
PDTCH PACCH
Page 5-19
1 2, 3 4 5 6, 7 8 9 10 11
The MS is assigned resources for the downlink transfer. The network sends the data blocks to the MS over the assigned downlink resources. The polling bit (S/P in the downlink MAC header) is set here. This requests the MS to acknowledge the data blocks previously received. The MS acknowledges the data blocks last received. Further blocks are transmitted to the MS. The network indicates a change in the downlink resources to the mobile and requests confirmation before further data is transmitted. The mobile acknowledges the changed downlink resources. Further downlink data block. This transfers the last data block of the current downlink TBF. The MS is requested (S/P) to acknowledge the end of the TBF. The MS recognizes the last data block by the Final Block Indicator in the downlink RLC header. The mobile acknowledges the end of the TBF. After this, the TFI is available for further transactions and the resources are released.
12
ACTERNA
MAC / RLC
Page 5-19
Packet Downlink ASSIGNMENT Message Packet Downlink ASSIGNMENT Message

< Packet Downlink Assignment message content > ::= < PAGE_MODE :bit (2) > {0 | 1 <PERSISTENCE_LEVEL :bit (4) >} { { 0 < Global TFI : < Global TFI IE > > |10 < TLLI : bit (32) > } { < MAC_MODE :bit (2) > < RLC_MODE :bit (1) > < CONTROL_ACK :bit (1) > < TIMESLOT_ALLOCATION :bit (8) > < Packet Timing Advance : < Packet Timing Advance IE > > {0 | 1 < P0 :bit (4) > < BTS_PWR_CTRL_MODE :bit (1) > } { {0 | 1 < Frequency Parameters : < Frequency Parameters IE > > } {0 | 1 < DOWNLINK_TFI_ASSIGNMENT :bit (5) > } {0 | 1 < Power Control Parameters : < Power Control Parameters IE > > } {0 | 1 < TBF Starting Time : < Starting Frame Number Description IE > > } {0 | 1 < Measurement Mapping : < Measurement Mapping structure > > }
Page 5-20
The figure shows and extract from the Packet Downlink Assignment message. The most important parameters are described below: MAC_MODE: Dynamic allocation or fixed allocation. RLC_MODE: Acknowledged or unacknowledged mode TIMESLOT_ALLOCATION: Downlink timeslot allocation. Maximum = 8. Frequency Parameters: Information about the radio frequency (ARFCN, frequency hopping, etc.). DOWNLINK_TFI_ASSIGNMENT: This assigns a unique TFI for the coming downlink TBF. TBF Starting Time: Absolute TDMA frame number from which the downlink transfer begins. Measurement Mapping: Contains information necessary for performing the measurements in its own and in neighboring cells. (Information about the timeslots to be used for level measurement, and information about the start of measurements and the measurement interval.)
ACTERNA
MAC / RLC
Page 5-20
Frequency Parameters Frequency Parameters
BTS
PACKET DOWNLINK ASSIGNMENT {frequency parameter}
PACKET UPLINK ASSIGNMENT {frequency parameter}
PACKET TIMESLOT RECONFIGURE {frequency parameter}
Page 5-21
Transfer of the radio frequency parameters is always in one of the three ASSIGNMENT messages shown above. The Frequency parameter information element contained in the messages defines a single frequency or a group of frequencies. It also contains information as to whether frequency hopping is to be performed or not.
ACTERNA
MAC / RLC
Page 5-21
Broadcast System Information Broadcast System Information

Everyone listens to my command ...
PBCCH PBCCH PSI-1 PSI-1 PSI-2 PSI-2 PSI-3 PSI-3 PSI-3bis PSI-3bis PSI4 PSI4 PSI5 PSI5
BCCH BCCH
PSI-13 PSI-13
PSI = Packet System Information
Page 5-22
If a PBCCH is available in the serving cell, the MS receives all the necessary system information (channel configurations, info on neighbor cells, measurement parameters, cell reselection criteria, etc.) via it. The PSI-13 info broadcast in the BCCH indicates whether a PBCCH is available. If there is no PBCCH in the current cell, the MS receives the system information via the BCCH. The elements PSI-1, -2, -3 are always present (mandatory elements). The optional elements PSI-3bis, -4, -5 can also be transmitted. As well as the broadcast channel, the network can also use the PACCH for transferring system information. This ensures that a specific mobile receives the information at least every 15 seconds.
ACTERNA
MAC / RLC
Page 5-22
The more important contents of the packet system information elements: PSI-1 Is the MS to transmit regular measurement reports or not Which type of access burst (11 bit or 8 bit) is to be used DRX yes / no Timeout value for PACKET ACCESS REQUEST Timeout value for PACKET CONTROL ACK/NACK Parameters for PRACH (number of access attempt repeats, minimum time between access procedures, priorities) PCCCH organization Global power control parameters PSI-2 Cell identification Cell allocation (frequencies of the current cell) GPRS mobile allocation (frequencies of the current cell for GPRS) Definitions for the PCCCH (ARFCN, timeslots, hopping) PSI-3 Serving cell parameters (minimum receive level for selecting the cell, max. MS transmit level in the PRACH, etc.) Parameters for cell reselection (wait time after unsuccessful cell reselection, level hysteresis, etc.) Parameters of the neighbor cells for cell reselection PSI-3bis Further neighbor cell parameters not already included in PSI-3. PSI-4 Information about interference measurements (on which PDCHs is the MS to perform interference measurements ?) PSI-5 All parameters necessary for controlling the measurement processes of an MS (on which frequencies / timeslots are measurements to be made, and at what intervals should a report be sent?). PSI-13 Access priority MS-controlled or network-controlled cell reselection? Description of PBCCH (if present)
ACTERNA
MAC / RLC
Page 5-23
Measurement Reports Measurement Reports
BTS
{Measurement parameter}
PACKET MEASUREMENT REPORT
Timer T3158
Page 5-24
Measurement parameter contents: NCO (MS-controlled cell reselection, no measurement reports) or NC1 (MS-controlled cell reselection, MS sends measurement reports) or NC2 (network-controlled cell reselection, MS sends measurement reports) Timer T3158 The measurement parameters are broadcast with PSI-5 or are selectively sent from the network to a particular mobile over the PACCH. The network uses the PACKET MEASUREMENT ORDER or PACKET CELL CHANGE ORDER messages for this. The measurement parameters contain all the information required by the MS for performing the level measurements.
Default values of Timer T3158: 3.8 s in packet transfer mode 61.44 s in packet idle mode
ACTERNA
MAC / RLC
Page 5-24
Cell Reselection Cell Reselection

Change yourself ...
Old BTS
New BTS
PACKET CELL CHANGE ORDER
PACKET CHANNEL REQUEST
PACKET UPLINK ASSIGNMENT
Example of a NC2 sequence
Page 5-25
In the case of independent cell reselection (NC0 or NC1), the MS continues any current transfers in the old cell until a successful change to the new cell takes place. Under no circumstances is transfer of data in the old cell to continue for more than 5 seconds after a change to a new cell. If necessary, the MS terminates every existing TBF immediately when a networkcontrolled cell change takes place. A new TBF is then set up after the change to the new cell. The message PACKET CELL CHANGE ORDER contains all the parameters needed to identify the new cell (e.g. BSIC, BCCH frequency, etc.). The MS considers a cell change order procedure as successfully completed as soon as it receives a positive response to its PACKET CHANNEL REQUEST message (the PACKET UPLINK ASSIGNMENT message in the above example).
ACTERNA
MAC / RLC
Page 5-25
Paging Procedure Paging Procedure

Hello, hello, ... Come in please
BTS
PACKET PAGING REQUEST
PACKET CHANNEL REQUEST
Page 5-26
The paging procedure with regard to a downlink packet data transfer can only be initiated by the network. The PACKET PAGING REQUEST message is transmitted in the PCCCH (subchannel PPCH) or in the CCCH. One or more mobile stations can be addressed simultaneously by each paging message.
ACTERNA
MAC / RLC
Page 5-26
PDCH Release PDCH Release
BTS
PACKET PDCH RELEASE
Page 5-27
The network transmits a PACKET RELEASE message in the PACH to indicate to an MS that certain timeslots are no longer available to it. If the release message does not contain a specific TIMESLOT_AVAILABLE field, the MS must immediately terminate its transmit and receive operations in the PDCH in which the release message was received, and remove this PDCH from its resource mapping table. Otherwise, all the PDCHs that are not specified in the TIMESLOT_AVAILABLE field should be treated as above, here, too, a (multislot-capable) mobile must terminate all transfers on unlisted PDCHs and remove them from its allocation list.
ACTERNA
MAC / RLC
Page 5-27
RLC/MAC Messages RLC/MAC Messages

Packet Access Reject Packet Access Reject Packet Channel Request Packet Channel Request Packet Queuing Notification Packet Queuing Notification Packet Resource Request Packet Resource Request Packet Uplink Assignment Packet Uplink Assignment Packet Downlink Assignment Packet Downlink Assignment Packet TBF Release Packet TBF Release Packet Paging Request Packet Paging Request Packet System Information Type 11 Packet System Information Type Packet System Information Type 22 Packet System Information Type Packet System Information Type 33 Packet System Information Type Packet System Information Type 3bis Packet System Information Type 3bis Packet System Information Type 44 Packet System Information Type Packet System Information Type 55 Packet System Information Type Packet System Information Type 13 Packet System Information Type 13 Other Other messages messages RLC RLC messages messages Packet Downlink ACK/NACK Packet Downlink ACK/NACK Packet Uplink ACK/NACK Packet Uplink ACK/NACK Packet Control Acknowledgement Packet Control Acknowledgement Packet Cell Change Failure Packet Cell Change Failure Packet Cell Change Order Packet Cell Change Order Packet Downlink Dummy Control Block Packet Downlink Dummy Control Block Packet Uplink Dummy Control Block Packet Uplink Dummy Control Block Packet Measurement Report Packet Measurement Report Packet Measurement Order Packet Measurement Order Packet Mobile TBF Status Packet Mobile TBF Status Packet PDCH Release Packet PDCH Release Packet Polling Request Packet Polling Request Packet Power Control/Timing Advance Packet Power Control/Timing Advance Packet PRACH Parameter Packet PRACH Parameter Packet Timeslot Reconfigure Packet Timeslot Reconfigure
Uplink TBF Uplink TBF establishment establishment
Downlink TBF Downlink TBF establishment establishment TBF release TBF release Paging Paging
System System information information
Page 5-28
All the messages mentioned have payload type = RLC/MAC control block. The length of an RLC/MAC control message is the same as the length of an RLC/MAC block or an integer multiple of it. To transmit an RLC/MAC signaling message, therefore, precisely n blocks (or 4 x n bursts) are required. Messages are filled with spare bits (padding) if necessary to bring them to the block limits. An RLC/MAC message usually contains the so-called error label. The receiver can derive how to handle an incorrect syntax message from this information (reject it or at least partially accept it).
ACTERNA
MAC / RLC
Page 5-28
Description of RLC/MAC signaling messages: There follows a broad overview describing the RLC/MAC signaling messages. The precise definitions, particularly the complete description of the contents (scope, mandatory or optional elements) are found in the ETSI Specification EN 301 349 (GSM 04.60). Packet Access Reject MS Network PCCCH or PACCH An Access Request from the MS is rejected by the network. Several mobiles can be addressed simultaneously by this message. Contents: Packet request reference TLLI (Temporary Logical Link Identifier) Minimum wait time before a new access attempt Packet Control Acknowledgement MS Network PACCH Response to certain messages from the network (e.g. Packet Polling Request or Packet Downlink Assignment, etc.). Acknowledges (positively) the message that was received. Contents: Acknowledgement TLLI (Temporary Logical Link Identifier) Packet Cell Change Failure MS Network PACCH The MS indicates to the network that a network-controlled cell change could not be performed successfully. Contents: TLLI (Temporary Logical Link Identifier) ARFCN of the new cell BSIC of the new cell CAUSE Packet Cell Change Order MS Network Order to the MS to change to a new cell. Contents: TFI ARFCN and BSIC of the new cell Measurement parameters
PCCCH or PACCH
ACTERNA
MAC / RLC
Page 5-29
Packet Channel Request MS Network PRACH Sent from the MS to the network in the form of an access burst in random mode in order to obtain resources e.g. for an uplink TBF. Contents: One-phase or two-phase request or Short Access Request Priority required Packet request reference (= TDMA frame number) Packet Downlink Ack/Nack MS Network PACCH Acknowledges receipt of RLC data blocks in Acknowledged RLC mode. Another use is to indicate the channel quality. Contents: Downlink TFI Ack/Nack description Channel description Channel quality report Packet Downlink Assignment MS Network PCCCH or PACCH The network assigns downlink resources to the MS with this message. Contents: Priorities Global TFI (uplink or downlink TFI to which the message refers) TLLI MAC mode (dynamic or fixed allocation) RLC mode (acknowledged or unacknowledged) Downlink TFI (that is assigned with this message) Timeslot allocation Timing advance Power control parameters Frequency parameters Measurement mapping Packet Downlink Dummy Control Block MS Network PCCCH or PACCH Filler message used if nothing else is to be transmitted. Required in order for the MSs to detect the USF in the MAC header. Contents: Page mode or Priorities (usually no content, however)
ACTERNA
MAC / RLC
Page 5-30
Packet Uplink Dummy Control Block MS Network Filler message sent by the MS if nothing else is to be transmitted. Contents: TLLI
PACCH
Packet Measurement Report MS Network PACCH Contains the measurement reports of the mobile station. Several messages of this type may be needed to transport all the results. Contents: NC mode (NC1 or NC2) Serving cell level values Serving cell interference level Neighbor cell frequencies Neighbor cell BSICs Neighbor cell level values Packet Measurement Order MS Network PCCCH or PACCH Contains orders to the MS regarding the performance of measurements and network-controlled cell reselection. Contents: Global TFI TLLI Measurement parameters Packet Mobile TBF Status MS Network PACCH The MS indicates to the network the receipt of a faulty message in connection with an uplink or downlink TBF. Contents: Global TFI TBF CAUSE Packet Paging Request MS Network PCCCH Sent simultaneously to up to 4 mobiles to trigger a channel request (e.g. in connection with a downlink block transfer). The MSs are addressed using their IMSI, TMSI or Packet TMSI. Contents: PTMSI (max. 4) IMSI (max. 4) TMSI (max. 4)
ACTERNA
MAC / RLC
Page 5-31
Packet PDCH Release MS Network PACCH The network indicates to all MSs that are monitoring this PDCH that one or more PDCHs must be released immediately and are no longer available for packet data transfer. Contents: Timeslot-available list (these PDCHs continue to be available) Packet Polling Request MS Network PCCCH or PACCH The network requests a Packet Control Acknowledgement message from the MS with this message (e.g. for monitoring the link). Contents: Global TFI TLLI Packet Power Control/Timing Advance MS Network PACCH The network informs the MS of the current values of timing advance or power control parameters. Contents: Global TFI Packet request reference Power control parameters Timing advance Packet PRACH Parameters MS Network PCCCH Sent to all MSs within one cell to inform them of an update of the PRACH parameters. Contents: PRACH control parameters Packet Queuing Notification MS Network PCCCH If the network cannot deal with a channel request from an MS immediately, it sends the affected MS this message informing the MS that its request has been placed on a waiting list (queue). Contents: Packet request reference Temporary queuing identifier (the MS is addressed with this as long as it is in the queue)
ACTERNA
MAC / RLC
Page 5-32
Packet Resource Request MS Network PACCH The MS requests the network to change the uplink resources. Contents: Global TFI TLLI MS radio access capability (multislot capability, ext. dynamic allocation, etc.) Channel request description Packet PSI Status MS Network PACCH The MS indicates to the network which Packet System Information message it has received. Contents: Global TFI Received PSI message list Packet System Information Type 1 MS Network PBCCH or PACCH Contains system-specific information about cell reselection, PRACH parameters, descriptions of control channels, and optional global power control parameters. Contents: Measurement order (NC0 yes / no) GPRS cell options (DRX timer, access burst type, etc.) PRACH control parameters PCCCH organization parameters Global power control parameters Packet System Information Type 2 MS Network PBCCH or PACCH Contains system-specific information about frequencies, cell allocation, PCCCH properties, etc. Contents: Cell identification (LAC, CI, RAC) Frequency lists Cell allocation GPRS mobile allocation PCCCH description (ARFCN, timeslot allocation, hopping sequence) Packet System Information Type 3 MS Network PBCCH or PACCH Contains system-specific information about neighbor cells and cell reselection parameters. Contents: Serving cell parameters (minimum receive level, max. output level @ packet control channel) General cell selection parameters (hysteresis, etc.) Neighbor cell parameters (number, frequencies, BSIC, minimum receive level, maximum output level @ PCCH, etc.)
ACTERNA
MAC / RLC
Page 5-33
Packet System Information Type 3bis MS Network PBCCH or PACCH Contains further system-specific information about neighbor cells that is not yet included in PSI-3. Contents: Neighbor cell parameters (number, frequencies, BSIC, minimum receive level, maximum output level @ PCCH, etc.) Network color code Packet System Information Type 4 MS Network PBCCH or PACCH Contains a list of the PDCHs on which the MSs should perform interference measurements during idle frames. Contents: Channel list for interference measurements (ARFCN, hopping, timeslots) Packet System Information Type 5 MS Network PBCCH Informs the MSs about the type and method of measurement reports as well as the network-controlled cell reselect procedures. Contents: Measurement parameters (frequencies, etc.) Network control order (NC0, NC1 or NC2) Report periods Packet System Information Type 13 MS Network BCCH or PACCH Normally sent in the BCCH but can also be sent in the PACCH to a specific mobile. If this message is sent in the BCCH it also contains the PBCCH mapping description if the PBCCH exists. Contents: Priority access threshold Network control order GPRS cell options GPRS power control parameters PBCCH description (training sequence, ARFCN, hopping, timeslot number) Packet TBF Release MS Network PACCH The network ends an uplink or downlink TBF and indicates this to the corresponding MS. Contents: Global TFI Uplink or downlink release TBF release cause
ACTERNA
MAC / RLC
Page 5-34
Packet Uplink Ack/NAck MS Network PACCH Serves as acknowledgement for received RLC data blocks in Acknowledged RLC mode. It is also used to transfer the timing advance and power control parameters. Contents: Uplink TFI Timing advance Power control parameters Packet Uplink Assignment MS Network PCCCH or PACCH The network assigns uplink resources to an MS with this message. The MS is addressed by TFI, TQI or packet request reference. Contents: Global TFI TQI Packet request reference Channel coding command (CS-1, CS2, ...) Timing advance Frequency parameters Type of allocation (fixed, dynamic, extended dynamic) Allocation parameters Packet Timeslot Reconfigure MS Network PACCH The network indicates the (changed) allocation of channels to an MS. Contents: Global TFI Channel coding command (CS-1, CS2, ...) Timing advance Downlink TFI assignment Uplink TFI assignment Frequency parameters Type of allocation Allocation parameters
ACTERNA
MAC / RLC
Page 5-35
Chapter 6 Chapter 6
Protocols Between BSS and SGSN Gb Interface
Page 6-1
Notes:
ACTERNA
Communications Between BSS and SGSN
Page 6-1

Application IP/X.25 SNDCP (or MM) LLC RLC RLC MAC GSM RF MAC GSM RF Relay BSSGP BSSGP Network service L1bis Network service L1bis IP L2 L1 IP L2 L1 Relay GTP SNDCP/MM LLC GTP UDP/ TCP UDP/ TCP IP/X.25
MS
Um
BSS
Gb
SGSN
Gn
GGSN
Gi
Page 6-2
The protocol stack is made up from the L1bis Layer, Network Service Layer and BSSGP Layer on the Gb interface between BSS and SGSN. L1bis: The physical Gb interface is not part of GPRS specifications, but must be agreed between operators and system manufacturers. This interface can take on very different forms. E1 or T1 interfaces are often found. To obtain the necessary bandwidth for user data (up to 1984 kbit/s), several timeslots must be used in parallel. Network Service Layer: The Network Service Layer is based as a rule on a Frame Relay connection (sub-network service). The following services are offered to the BSSGP layer: Transport of NS SDU (NS Service Data Units) Indication of possible network overload conditions Indication of status information BSSGP Layer: The BSSGP (Base Station Subsystem GPRS Protocol) mainly takes care of the following tasks: Providing connectionless communications between the SGSN and BSS Transport of unacknowledged data blocks between the SGSN and BSS Bi-directional flow control Handling paging requests Deleting old BSS message buffers (flush function) Supporting multiple parallel links between the BSS and SGSN
ACTERNA
Page 6-2
Network Service: Internal Architecture Network Service: Internal Architecture

Network service Network service
Network service control Network service control

Network service control protocol Network service control protocol
Sub-network service Sub-network service

Sub-network service protocol Sub-network service protocol
e.g. Frame Relay
Page 6-3
The Network Service Layer provides its user (the BSSGP) with peer-to-peer communications facilities. Communication between distant Network Service control instances is via Network Service Virtual Connections (NS-VCs). An NS-VC is thus a virtual path between NS control instances. As well as transporting the user data, the Network Service Control Sub-layer is also responsible for load-sharing over the available NS-VCs and for VC management. The type and manner of technical realization of the Sub-Network Service Sub-Layer is not specifically fixed in the GPRS specifications, but recommendations are made for direct link configurations (e.g. PP fixed connections) and Frame Relay networks.
ACTERNA
Page 6-3
Network Service: Addressing Network Service: Addressing

End to end NS-VC
NS-VL (BSS side)
NS-VL (SGSN side)
Transmission network
BSS BSS (e.g. Frame Relay) SGSN SGSN
Gb interface
Page 6-4
The diagram shows the basic addressing principle on the Gb interface. Network Service Virtual Link (NS-VL) SGSN and BSS can be connected together via different physical links. This is the case if there is a transport network between the two network nodes. Each physical link is locally addressed using different physical link identifiers. The exact link structure depends on the implementation. Every physical link can accommodate one or more NS-VLs. The precise appearance of the NS-VLs is also dependent on the implementation. Every NS-VL is identified by a Network Service Link Identifier (NS-VLI). Network Service Virtual Connection (NS-VC) The virtual connection concept was devised in order to provide end-to-end communication between the BSS and SGSN regardless of the Gb interface configuration. In the case of a Frame Relay network, the FR Permanent Virtual Connection (PVC) corresponds to the NS-VC. Network Service Virtual Connection Group Several virtual connections between BSS and SGSN are collected together into a group. BSSGP Virtual Connection (BVC) The Network Service Layer provides communication paths between remote NS users (BSSGP instances). These paths are called BSSGP Virtual Connections and are based on Network Service Virtual Connections.
ACTERNA
Page 6-4
Network Service: Load Sharing Network Service: Load Sharing
NS-VC group
NS-VC 1 NS-VC 2 NS-VC 3 NS-VC 4 NS-VC n
BSS BSS
SGSN SGSN
Gb interface
Page 6-5
Transport of NS Service Data Units (NS-SDU) is evenly distributed over the available virtual connections in a group of NS-VCs. The order in which the NS-SDUs are received at the receiver will very likely be different to the order in which they were transmitted, however. Distribution of the data over different NS-VCs is done using a Link Selector Parameter (LSP). A fixed LSP leads to transmission of all data on (only) one defined NS-VC. In this case, the transmit and receive order will be the same.
ACTERNA
Page 6-5
Network Service: VC Management Functions Network Service: VC Management Functions
Blocking and unblocking NS-VCs
Reset a NS-VC
Test a NS-VC
Page 6-6
Blocking / Unblocking function If the unavailability of a VC is established by the BSS or SGSN, the VC is indicated as blocked and the opposite end informed of this by a blocking procedure. The opposite end then also takes the affected VC out of service. Blocking of a VC can be triggered by: Maintenance BSS or SGSN malfunction Link errors Faults in the transit network (e.g. faults in the Frame Relay network) etc. A VC can be put back into service as soon as it is available again. An unblocking procedure is executed for this, during which the communication partners indicate the new availability and then mark the VC as unblocked. Reset function Forces a defined state for a VC on both sides (e.g. after re-commissioning a VC or after a processor restart). Test function Checks the function of a VC. The result is either alive or dead.
ACTERNA
Page 6-6
Network Service: Service Primitives Network Service: Service Primitives

Network service user (BSSGP) Network service user (BSSGP)
Request
Indication
Request: UNITDATA {BVCI, NSEI, NS-SDU, LSP}
Indication: UNITDATA {BVCI, NSEI, NS-SDU} CONGESTION {BVCI, NSEI, Congestion cause} STATUS {BVCI, NSEI, NS affecting Cause, Transfer capability
Page 6-7
All BSSGP data are transmitted as UNITDATA elements between the layers. These are sent as Requests on the transmit side and as Indications on the receive side. Information about overloads and status is sent to the BSSGP in the form of an indication (CONGESTION or STATUS). The parameters have the following meanings: BVCI (BSSGP Virtual Connection Identifier) Identifies a logical virtual connection between BSS and SGSN on the BSSGP layer. NSEI (Network Service Entity Identifier) Identifies a group of several NS-VCs. LSP (Link Selector Parameter) This parameter is used to select a particular NS-VC from within a group of NS-VCs. The group is selected by the NSEI. NS-SDU (NS Service Data Unit) Contains the user data of the BSSGP. NS-SDUs are transported transparently by the Network Service. Congestion Cause Contains information about congestion and its direction (forward congestion or backward congestion). Transfer Capability Contains information about the currently available bandwidth (corresponding to the number of NS-VCs used). NS Affecting Cause Contains information about the cause of the status indication
ACTERNA
Page 6-7
Network Service: Sub-Network Network Service: Sub-Network
BSS BSS (User) (User)
Gb
SGSN SGSN (Network) (Network)
Direct link configuration
BSS BSS (User) (User)
Gb Frame Relay network
Gb
SGSN SGSN (User) (User)
PVC via Frame Relay network
Page 6-8
The Gb interface can either be a point-to-point connection between the BSS and SGSN (1) or permanent virtual connections over an interposed Frame Relay network are used (2).
The following assignments apply for a Frame Relay network: Physical link NS-VL NS-VLI NS-VC = Frame Relay Bearer Channel = User Network Interface (UNI) = DLCI = Permanent Virtual Connection (PVC)
ACTERNA
Page 6-8
Network Service: Frame Relay Network Service: Frame Relay

Features of Frame Relay:

Economical (minimal resource) transmission of data packets High transmission speed Connection-oriented protocol (similar to HDLC) in Layer 2 of the OSI model Only a few, simple security procedures (therefore fast) Dedicated lines used as tributaries to the Frame Relay nodes Typical interfaces: X.21, V.35 and G.703/G.704 at 64 kbit/s to 2 Mbit/s. PMP connections also possible as well as PP Guaranteed minimum bandwidth High data throughput Low network delay Segmentation hardly necessary as packet size is large Network can be easily upgraded Protocol-transparent Simple network administration
Page 6-9
HDLC frames are used for data transfer in FR. High-performance lines (optical fibers, etc.) require few protection procedures as the bit error rates are low. FR only provides error detection, and no error correction. A higher data throughput is achieved in the FR network by doing without error correction and flow control. Error correction must take place in higher protocol layers in the terminal equipment. FR provides users with a guaranteed minimum bandwidth. A committed information rate (CIR) can be agreed for each logical connection. The CIR is always available to the FR user, but is always less than the connection speed. The FR user can also feed data into the FR network at rates higher than the CIR, but such data will be destroyed by the network if there is congestion.
ACTERNA
Page 6-9
Network Service: Frame Relay Interface Network Service: Frame Relay Interface
FR node
DCE
DCE
DTE UNI
Frame Relay network
DTE
UNI
Page 6-10
UNI DTE DCE
= User to Network Interface = Data Terminal Equipment = Data Communications Equipment
Every DTE (FR client) can support one or more permanent logical connections at its UNI with different guaranteed bandwidths to any other DTE. The logical PVCs are fixed in the network by the FR network operator (by configuring the routing tables in each of the participating FR nodes).
ACTERNA
Page 6-10
Network Service: Frame Relay Frame Structure Network Service: Frame Relay Frame Structure
Flag (7EH) Flag (7EH)
Header Header
Data Data
FCS FCS
DLCI DLCI
FECN BECN DE FECN BECN DE
DLCI = Data link connection identifier FECN = Forward explicit congestion notification BECN = Backward explicit congestion notification DE = Discard eligibility indicator
Page 6-11
DLCI (17 or 23 bits) The Data Link Connection Identifier identifies a permanent virtual connection (PVC). It is often called the address that describes the path through the FR network. FECN / BECN (1 bit each) The FECN and BECN bits indicate congestion in the network to the terminal equipment. They also indicate the direction in which the network is congested. DE (1 bit) If a FR network is congested, it destroys the frames that cannot be processed. Frames with DE=1 are destroyed before frames with DE=0. BSS and SGSN always set DE to 0. Flag (1 byte) Used to identify the start and end of the frame. The flags are binary coded with 01111110. This pattern is avoided between flags by means of the bit stuffing mechanism. FCS Frame Checking Sequence. Used for detecting transmission errors. The FCS corresponds to a division of the frame (header + data) by the generator polynomial x16 + x12 + x5 + 1. Data The data field is used to transport transparent user data from higher layers. The length can be several thousand bytes. The data field for GPRS is a maximum of 1600 bytes long. Greater lengths are possible but have to be agreed separately between the GPRS and FR operators.
ACTERNA
Page 6-11
Network Service: Frame Relay Routing Network Service: Frame Relay Routing
DLCI 52 DLCI 18
DLCI 33
DLCI 21
TE B TE B TE A TE A
DLCI 45 DLCI 57
Frame Relay switch

DLCI 21
TE C TE C
Page 6-12
Frame Relay uses permanent virtual connections (PVCs). The Data Link Connection Identifier (DLCI) is used to address these connections. The PVCs must be permanently installed by the network operator by means of a network management system. Several virtual connections can exist simultaneously at a connection port (UNI). A DLCI only has local meaning in local addressing. The DLCI takes on different values in the various segments on the path through the network. Every logical connection on the connecting circuit is described by exactly one DLCI. In the above example TE A has the DLCI values 33 and 45.
ACTERNA
Page 6-12
Network Service: Frame Relay Parameters Network Service: Frame Relay Parameters
Data volume
Access rate Frames destroyed Bc + Be Flag frames with DE bit
Bc
CIR
Transmission guaranteed
T0 Measurement time Tc
T0 + Tc
t/s
Page 6-13
Bc = Committed Burst Size Be = Excess Burst Size Tc = Measurement Interval The parameters CIR, Bc, Be and Tc each apply to a single permanent virtual connection. The decisive parameter in a FR connection is the Committed Information Rate (CIR). It is usually less than the physical connection speed of the connecting circuit (e.g. below 2 Mbit/s for G.703/G.704). Bc determines the agreed quantity of data that may be placed on the FR network by the user within the time interval Tc and which will be transported by the network to the destination transparently. Be specifies the excess data quantity that can be put on to the network in the same time interval without the data being rejected at the network input. The packets will, however be indicated with the DE bit = 1. Frames in excess of the limit of Bc + Bc during Tc will be rejected by the input node directly.
ACTERNA
Page 6-13
Network Service: Frame Relay Signaling Network Service: Frame Relay Signaling
DTE
STATUS ENQUIRY {DLCI = 0}
DCE
Terminal Terminal equipment equipment
FR node FR node STATUS {DLCI = 0}
FR network
Page 6-14
It has proven useful in operating Frame relay networks and terminal equipment to exchange information between the network and the attached terminal equipment in addition to the pure data transfer. This exchange of information is designated as Local Management Interface (LMI) and in-channel signaling in FR and gives information about: connection circuit status newly set up PVCs deleted PVCs active or inactive operating states of the individual PVCs
Signaling is based ona periodic exchange of messages between the DTE and DCE. The DTE sends status queries to the DCE which then responds with a status report. The DLCI is set to 0 for in-channel signaling.
ACTERNA
Page 6-14
Network Service: Control Protocol Network Service: Control Protocol

The NS control protocol allows the following functions:

Transmission of BSSGP user data (NS-SDUs) Blocking and unblocking procedures Reset procedure Test procedure Error report procedures
VC management functions
Page 6-15
NS-SDUs are transmitted in both directions in unacknowledged mode over the Gb interface. The error report procedure informs opposite protocol instances of protocol errors that occur. The error information I transported by the NS status PDU (Protocol Data Unit).
ACTERNA
Page 6-15
Network Service: --Encapsulation Network Service: Encapsulation
1 byte PDU type
Max. 1599 bytes Information elements
Header Header
FR data frame FR data frame
FCS FCS
Frame Relay transmission frame
Page 6-16
The messages of the NS Control protocol are called Protocol Data Units (PDUs). The PDUs contain information about the PDU type along with further mandatory or optional information elements. All together, a PDU can have a maximum length of 1600 bytes. It is transferred transparently in the data field of the FR frame.
ACTERNA
Page 6-16
Network Service: Control PDUs Network Service: Control PDUs

PDU type NS-ALIVE NS-ALIVE-ACK NS-BLOCK NS-BLOCK-ACK NS-RESET NS-RESET-ACK NS-STATUS NS-UNBLOCK NS-UNBLOCK-ACK NS-UNITDATA Direction BSS SGSN Use
Testing a VC Response to NS-ALIVE Request to other end to block a VC Response to NS-BLOCK The transmitter of this PDU indicates a reset of a VC Acknowledgement of reset requested by NS-RESET Error report to the opposite end Request to unblock a previously blocked VC Acknowledgement of a NS-UNBLOCK message Transport element for NS-SDUs (BSSGP user data)

Page 6-17
NS-UNITDATA is the only PDU used for transporting BSSGP messages. All other PDU types are used exclusively for signaling in the Network Service layer.
ACTERNA
Page 6-17
PDU type NS-ALIVE NS-ALIVE-ACK NS-BLOCK
Parameter (IE) Cause NS-VCI
M/O
M M M M M M M M M O *) O O
NS-BLOCK-ACK
NS-VCI Cause
NS-RESET
NS-VCI NSEI NS-VCI
NS-RESET-ACK NSEI Cause NS-VCI NS-STATUS NS-PDU BVCI NS-UNBLOCK NS-UNBLOCK-ACK NS-UNITDATA NS-SDU *) M BVCI M
M = Mandatory O = Optional (linked to conditions)

*) General information: PDUs (Protocol Data Unit) transport data on a layer level, e.g. control or status information between two protocol automats in the same layer. SDUs (Service Data Unit) contain data from higher protocol layers, which is encapsulated and transmitted transparently by this method. The contents of the SDU are transferred to the higher protocol layer again in the receiver.
ACTERNA
Page 6-18
Trace Example 1 Trace Example 1
Page 6-19
This example shows a broad overview of the messages recorded at the Gb interface in conjunction with a Frame Relay transmission network. . BSSGP messages are transported as NS-UDT (Network Service Unit Data) and contain a BSSGP Virtual Connection Identifier (BVCI) to distinguish between several possible parallel logical connections in the BSSGP layer. Messages that are only exchanged in the Network Service layer do not contain a BVCI. Addressing only uses the DLCI in this case.
ACTERNA
Page 6-19
Page 6-20
Detail interpretation of an NS-ALIVE PDU
ACTERNA
Page 6-20
Page 6-21
Detail interpretation of an NS-STATUS PDU
ACTERNA
Page 6-21
BSSGP Service Model BSSGP Service Model
Relay Relay
GMM GMM
NM NM
LLC LLC
GMM GMM
NM NM
RL
GMM
NM
BSSGP
GMM
NM
RLC/MAC RLC/MAC
BSSGP BSSGP
NS SAP
BSSGP BSSGP
NS SAP
Service model in the BSS
Service model in the SGSN
Page 6-22
The Service Model is not symmetrical between the BSS and SGSN. For example, LLC connections pass through the BSS transparently and terminate in the MS. The following processes or protocols are BSSGP (Base Station Subsystem GPRS Protocol) users on the BSS side: The relay process, which takes care of matching between the BSSGP and RLC/MAC on the radio interface. GPRS Mobility Management (GMM) Network Management (NM) The following protocols / processes use the services of the BSSGP on the SGSN side: Logical Link Control (LLC) GPRS Mobility Management (GMM) Network Management (NM) There is a one-to-one relationship between the BSS and the SGSN, i.e. each BSSGP layer serves only one Gb interface. If a SGSN is connected to several BSS, the SGSN requires a separate BSSGP automat for each BSS that is connected.
ACTERNA
Page 6-22
The BVCI, LSP and NSEI Routing Parameters The BVCI, LSP and NSEI Routing Parameters
BVCI BVC 1 BSSGP BSSGP BVC n BSSGP automat 1 BSSGP BSSGP BSSGP automat 2 BVC n BVCI BVC 1 BSSGP BSSGP
NSEI LSP VC 1 Network service Network service VC n Network Service Network Service NS entity 1 NS entity 2 VC n LSP VC 1 Network service Network service
BSS
SGSN
Group of virtual connections
BSS
Page 6-23
The parameters BVCI, NSEI and LSP are used to address and route data packets and signaling packets between the SGSN and BSS. BVCI (BSSGP Virtual Connection Identifier) Identifies a BVC in the BSSGP layer. The BVCI can occur more than once in an SGSN but must be unique with reference to a Network Service Entity. In fact, the BVCI (value > 1) addresses a radio cell within a BSS for the purpose of user data transfer. BVCI = 0 reserved for BSSGP signaling BVCI = 1 Point-to-multipoint connection BVCI > 1 Data packets for point-to-point transfer NSEI (Network Service Entity Identifier) As a rule there are several Network Service entities in the NS layer at a network node. These NS entities are addressed with the NSEI parameter. The NSEI also simultaneously addresses a group of virtual connections between the NS protocol instances on the BSS and SGSN side, i.e. every NS entity is assigned a group of VCs. LSP (Link Selector Parameter) This parameter is used to address a particular VC in a group of VCs. The mapping between the LSP and the actual VC strongly depends on the implementation. If the LSP is fixed, all the data will be transferred over one and the same VC. If different LSPs are used for each data frame, they will be spread across the available VCs (load sharing).
ACTERNA
Page 6-23
Services on RL and BSSGP SAPs Services on RL and BSSGP SAPs

The relay process in the BSS and the LLC process in the SGSN can access the following BSSGP services:

Transmission of transparent downlink data packets
Transmission of transparent uplink data packets
Transmission of transparent point to multipoint downlink data packets
Page 6-24
Notes:
ACTERNA
Page 6-24
DL/UL Transfer on RL and BSSGP SAPs DL/UL Transfer on RL and BSSGP SAPs
BSS
Relay process Relay process
SGSN
LLC LLC
REQUEST { BVCI, NSEI, LSP, UL-UNITDATA-PDU }
SAP RL
INDICATION { BVCI, NSEI, DL-UNITDATA-PDU }
REQUEST { BVCI, NSEI, LSP, DL-UNITDATA-PDU }
SAP BSSGP
INDICATION { BVCI, NSEI, UL-UNITDATA-PDU }
BSSGP BSSGP Network service Network service
Downlink transfer Uplink transfer
Page 6-25
Transfer of downlink data packets takes place at the SAPs as so-called DLUNITDATA PDUs (downlink) or UL-UNITDATA PDUs (uplink). In addition to the user data, these PDUs contain all the other parameters necessary to establish a peer-topeer connection between the network and the mobile and to address the mobile (e.g. TLLI, IMSI, etc.). The following parameters are transferred at the RL and BSSGP SAPs in addition to the UNITDATA PDUs: BVCI NSEI LSP So-called PTM-UNITDATA PDUs are defined in the standard in addition to the uplink and downlink UNITDATA PDUs. These are transmitted between the layers in the case of a point-to-multipoint connection where the same data packet is to be sent in the downlink to several mobiles simultaneously.
ACTERNA
Page 6-25
Services on SAP GMM Services on SAP GMM

The following BSSGP services are available for GPRS mobility management:

Paging procedure Radio access capability update procedure Radio status procedure Suspend and resume procedure
Page 6-26
Paging procedure This procedure is used to page a mobiles within a BSS area. Radio Access Capability Update procedure A BSS sends a query to its SGSN regarding the current Radio Access Capability (multislot capability, channel allocation possibilities, etc.) of a particular mobile and receives the desired information from the SGSN in the form of an RA CAPABILITY UPDATE ACK PDU, which also contains the current TLLI of the MS. Radio Status procedure A radio link between the BSS and MS can fail accidentally or due to some error. This can happen with: problems in radio coverage unsuccessful cell reselection poor transmission quality In all these situations, the BSS indicates the cause of the connection failure to its SGSN in the form of a RADIO STATUS PDU. Suspend and Resume procedure If a MS wants to suspend the GPRS connection temporarily, it indicates this to the BSS, which then sends a SUSPEND PDU to the SGSN. Resumption of the GPRS connection by the MS is indicated to the SGSN by a RESUME PDU.
ACTERNA
Page 6-26
Paging Procedure on SAP GMM Paging Procedure on SAP GMM

BSS
GPRS GPRS mobility management mobility management
SGSN
GPRS GPRS mobility management mobility management
SAP GMM
INDICATION { BVCI, NSEI, PAGING-PS-PDU or PAGING-CS-PDU }
REQUEST { BVCI, NSEI, PAGING-PS-PDU or PAGING-CS-PDU }
Page 6-27
To perform a paging procedure, a SGSN sends one or more Paging PDUs to a BSS. These PDUs contain all the information necessary for locating a mobile within a cell or group of cells. There is a difference between GPRS paging (PS = Packet Switched) and GSM paging (CS = Circuit Switched). The following paging resolutions are possible: a single cell (i.e. one BVCI) all cells of the BSS all cells in a location area all cells in a routing area To determine the paging group, the BSS also receives the IMSI and, optionally, the DRX parameters of the mobile being paged. A PAGING PS PDU contains the following parameters: IMSI DRX parameters BVCI or Location Area or Routing Area or BSS Area Indication QoS Profile P-TMSI
ACTERNA
Page 6-27
Services on SAP NM Services on SAP NM

Network management is based on the following BSSGP services:

Flow control procedure Flush procedure BVC blocking and unblocking procedure BVC reset procedure Trace procedure
Page 6-28
Flow Control Since transfer of data packets to the MS is controlled by the BSS, the flow control function must have a downlink buffer for every BVC from the point of view of the BSSGP. These downlink buffers are BSS elements. Flow control is only implemented in the downlink. Flush procedure The SSGN sends a FLUSH LL PDU to the BSS affected by an MS cell change. The parameters in this PDU are the old BVCI (the old cell) and, optionally, the new BVCI (the new cell) together with the current TLLI for addressing the MS. The BSS then transfers the buffered data that was not yet sent to the MS from the old to the new BVCI. If a new BVCI is not specified, the BSS clears the buffer memory of the old BVCI. BSSGP signaling (BVCI=0) is excluded from the flush procedure. BVC Blocking and Unblocking This procedure is for blocking and unblocking individual BVCs. Possible causes that require blocking of a BVC: O&M on a cell System error Fault within a radio cell, etc. The blocking function is not possible for BSSGP signaling (BVCI=0).
ACTERNA
Page 6-28
BVC Reset This function sets the BVC and the corresponding context between the BSS and SGSN to a defined initial state and re-synchronizes both sides on the BSSGP layer. The procedure can occur in connection with: system errors in the BSS or SGSN system errors within a Network Service layer changes in BVCI / cell mapping etc.
Trace procedure A trace can be started from the OMC to monitor an individual mobile or to locate a fault. This takes place in the BSSGP layer by the SGSN sending a SGSN INVOKE TRACE PDU to the BSS. The parameters of this PDU are Trace Type, Mobile ID, OMC ID, and others.
ACTERNA
Page 6-29
Frame Structure of the Protocol Messages Frame Structure of the Protocol Messages
7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0
0x7E DLCI DLCI

C/R 0 FE BE DE CN CN 1
NS-UNITDATA BVCI (0)
PAGING PS PDU IMSI
FR user data (max. 1600 bytes) NS-SDU
DRX parameters
BVCI (>1)
FCS 0x7E NS sub-network service protocol Network service control protocol
QoS profile
BSSGP
Page 6-30
Example of a Packet Paging message to demonstrate message encapsulation.
The BSSGP instance in the SGSN receives a Paging Request from Mobility Management. The BSSGP process hands on this data to the subordinate Network Service process in the form of a PAGING PS PDU. The data is encapsulated there into an NS UNITDATA PDU and again passed on to the subordinate Sub-Network Service instance (a Frame Relay function in this example). Packed in a FR frame, the message reaches the Gb interface and can be recorded and displayed in this form using a protocol analyzer. The BVCI in the PAGING PDU points to the cell in the BSS where the MS is to be looked for. The BVCI in the NS UNITDATA PDU has the value 0, since this is a BSSGP signaling message. Signaling is always transmitted on BVCI 0, in contrast with user data. In other words, a BVCI > 0 terminates in a cell and the BVCI = 0 terminates in the BSSGP process of the BSS.
ACTERNA
Page 6-30
Start of NS-UNITDATA PDU
Start of UL-UNITDATA PDU
Page 6-31
Detail interpretation of trace example.
ACTERNA
Page 6-31
PDU Definitions PDU Definitions

DL-UNITDATA PDU (SAP RL and BSSGP) Information element PDU type: DL-UNITDATA Current TLLI QoS profile PDU lifetime MS radio capability Priority DRX parameters IMSI Old TLLI Alignment octets LSA information LLC-PDU M/O M M M M O O O O O O O M Length / Bytes 1 4 3 4 7-15 3 4 5-10 6 2-5 7-? 3-? M = Mandatory, O = Optional PAGING PS PDU (SAP GMM) Information element PDU type: PAGING-PS IMSI DRX parameters BVCI *) Location area *) Routing area *) BSS area indication *) QoS profile P-TMSI M/O M M O O O O O M O Length / Bytes 1 5-10 4 4 7 8 3 5 6
*) One and only one of these parameters is present
Page 6-32
Overview of all PDUs:

[Parameter name] = Mandatory parameters, M {Parameter name} = Optional parameters, O <Parameter name> = Conditional, C (mandatory element dependent on context)
PDUs at the RL and BSSGP SAP DL-UNITDATA [PDU type, Current TLLI, QoS profile, PDU lifetime, LLC-PDU] {MS Radio Access Capability, Priority, DRX parameters, IMSI, Old TLLI, Alignment octets, LSA information} DL For transporting a LLC packet to the mobile over the radio interface. [PDU type, TLLI, QoS profile, Cell Identifier, LLC-PDU] {Alignment octets, LSA identifier list} UL For transporting a LLC packet from the mobile to the SGSN.
Direction: Application: UL-UNITDATA Direction: Application:
RA-CAPABILITY [PDU type, TLLI, MS Radio Access Capability] Direction: DL Application: Informs the BSS about a new radio capability of the mobile.
ACTERNA
Page 6-32
PDUs at the GMM SAP
PAGING PS
Direction: Application:
[PDU type, IMSI, QoS profile] {DRX parameters, P-TMSI} <BVCI, Location area, Routing area, BSS area indication> DL Packet paging an MS for a GPRS downlink connection.
PAGING CS
[PDU type, IMSI, DRX parameters] {TLLI, Channel needed, eMLPP priority, TMSI} <BVCI, Location area, Routing area, BSS area indication> DL Paging for a circuit-switched GSM connection.
RA-CAPABILITY_UPDATE Direction: Application:
[PDU type, TLLI, Tag] UL Query for a current radio access capability
RA-CAPABILITY-UPDATE-ACK Direction: Application:
[PDU type, TLLI, Tag, Cause] <IMSI, MS Radio Access Capability> DL For indicating the current radio access capability.
RADIO-STATUS Direction: Application:
[PDU type, Cause] <TLLI, TMSI, IMSI> UL The BSS informs the SGSN about an exceptional state on the radio interface.
SUSPEND Direction: Application:
[PDU type, TLLI, Routing area] UL Indicates that an MS wishes to suspend GPRS traffic temporarily.
SUSPEND-ACK Direction: Application:
[PDU type, TLLI, Routing area, Suspend reference number] DL Positive response to receipt of SUSPEND
SUSPEND-NACK Direction: Application:
[PDU type, TLLI, Routing area] {Cause} DL Negative response to SUSPEND
ACTERNA
Page 6-33
RESUME Direction: Application: RESUME-ACK Direction: Application:
[PDU type, TLLI, Routing area, Suspend reference number] UL The MS indicates the resumption of a suspended connection. [PDU type, TLLI, Routing area] DL Positive response to RESUME.
PDUs at the NM SAP FLUSH-LL Direction: Application: [PDU type, TLLI, BVCI old] {BVCI new} DL This PDU informs the BSS that an MS has moved to another cell. The BSS then performs the flush function.
FLUSH-LL-ACK [PDU type, TLLI, Flush action, Number of octets affected] <BVCI new> Direction: UL Application: Response to FLUSH-LL. The BSS indicates the type of flush action (deletion or transfer to new BVCI) and the number of bytes affected to the SGSN. LLC-DISCARDED Direction: Application: [PDU type, TLLI, LLC frames discarded, BVCI, No. of octets deleted] UL Indicates that LLC PDUs buffered within a cell were deleted e.g. because their lifetime had expired. [PDU type, Tag, BVC bucket size, Bucket leak rate, Bmax default MS, R default MS] {BVC measurement} UL Informs the flow control process in the SGSN of the status and the current maximum acceptable throughput of a BVC. [PDU type, Tag] DL Acknowledgement for a FLOW-CONTROL-BVC PDU.
FLOW-CONTROL-BVC
FLOW-CONTROL-BVC-ACK Direction: Application: FLOW-CONTROL-MS Direction: Application:
[PDU type, TLLI, Tag, MS bucket size, Bucket leak rate] UL Informs the flow control process in the SGSN of the status and the current maximum acceptable throughput of an MS.
ACTERNA
Page 6-34
FLOW-CONTROL-MS-ACK Direction: Application: BVC-BLOCK Direction: Application:
[PDU type, TLLI, Tag] DL Acknowledgement of a FLOW-CONTROL-MS PDU
[PDU type, BVCI, Cause] UL The receiver of this PDU is to block the corresponding BVC. [PDU type, BVCI] DL Acknowledgement of a BVC-BLOCK PDU. [PDU-Type, BVCI] UL The SGSN is to unblock the specified BVC for further use. [PDU type, BVCI] DL Acknowledgement of an unblocked BVC.
BVC-BLOCK-ACK Direction: Application: BVC-UNBLOCK Direction: Application: BVC-UNBLOCK-ACK Direction: Application: BVC-RESET Direction: Application:
[PDU type, BVCI, Cause] <Cell identifier> DL, UL Request to reset a BVC. [PDU type, BVCI] <Cell identifier> DL, UL Acknowledgement of a BVC reset.
BVC-RESET-ACK Direction: Application: STATUS
Direction Application:
[PDU type, Cause] {PDU with error} <BVCI> DL, UL Indicates an exception state (e.g. receipt of an invalid PDU). [PDU type, Trace type, Trace reference] {Trigger ID, Mobile ID, OMC ID, Transaction ID} DL The SSGN orders the BSS to activate a trace.
SGSN-INVOKE-TRACE Direction: Application:
ACTERNA
Page 6-35
Chapter 7 Chapter 7
Page 7-1
Notes:
ACTERNA
Page 7-1

Application IP/X.25 Relay L3 protocols L3 prot. LLC RLC RLC MAC GSM RF MAC GSM RF Relay BSSGP BSSGP Network service L1bis Network service L1bis IP L2 L1 IP L2 L1 LLC GTP UDP/ TCP UDP/ TCP GTP IP/X.25
MS
Um
BSS
Gb
SGSN
Gn
GGSN
Gi
Page 7-2
Notes:
ACTERNA
Page 7-2
LLC
(Logical Link Control Protocol)
Page 7-3
The LLC layer can be considered as a sub-layer of layer 2 of the OSI reference model. The task of the LLC layer is error-free transport of the layer 3 data between the MS and SGSN. To do this, LLC provides high-reliability logical connections between the MS and SGSN. LLC is independent of the radio layer below it. This allows physical transmission variants (UMTS) to be used in future.
ACTERNA
Page 7-3
LLC Layer Reference Model LLC Layer Reference Model
TOM
GMM
SNDCP
SMS
L3 protocols
TOM
GMM
SNDCP
SMS
LLC RLC RLC MAC GSM RF MAC GSM RF Relay
LLC
BSSGP BSSGP Network service L1 s Service L1
MS
Um
BSS
Gb
SGSN
Page 7-4
The LLC layer connects the MS and SGSN transparently across the BSS. LLC serves as the transport protocol for the L3 layer data packets (Protocol Data Units, PDUs in OSI terminology). Transport in the LLC layer is through logical channels that are separated by the Data Link Connection Identifier (DLCI). LLC: TOM: GMM: SNDCP: SMS: Logical Link Control Tunneling of Messages GPRS Mobility Management Sub-network Dependent Convergence Protocol Short Message Service
ACTERNA
Page 7-4
LLC Features LLC Features
Support for several MSs simultaneously on the Um interface. Support for several simultaneous L3 processes within a MS. LLC provides a separate logical connection. Monitoring the order of data frames Error detection Error correction Indication of uncorrectable errors Flow control Data ciphering
Page 7-5
As well as the features described above, the following general requirements apply to the LLC layer: Providing highly reliable logical connections Independence from the radio interface Transport frame length should be variable Support for peer-to-peer connections Provision for acknowledged and unacknowledged transmission Multiplexing of different MS on to a common physical channel Providing priority mechanisms during transmission Preserving user data confidentiality (data ciphering) Preserving subscriber identity confidentiality (via the TLLI)
ACTERNA
Page 7-5
Acknowledged //Unacknowledged Modes Acknowledged Unacknowledged Modes

MS/ MS/ SGSN SGSN
SABM UA I frame {N(S), N(R)} RR {N(R)} I frame {N(S), N(R)} RR {N(R)} DISC UA UI frame UI frame UI frame
MS/ MS/ SGSN SGSN
MS/ MS/ SGSN SGSN

UI frame
MS/ MS/ SGSN SGSN
Acknowledged mode
Unacknowledged mode
Page 7-6
There is a choice of two basic transport mechanisms available in the LLC layer: 1. Acknowledged Mode The PDUs of higher layers are transported in numbered I frames. The received I frames are acknowledged in the LLC layer, e.g. with Receive Ready frames (RR). Error detection and correction based on repetition of frames is implemented. In this mode, flow control is possible. Acknowledged mode is initialized by exchanging the messages SABM and UA. To end Acknowledged mode, the processor resources are released by exchanging DISC and UA messages. 2. Unacknowledged Mode: The PDUs of the higher layers are transported in unnumbered (UI) frames. No acknowledgement takes place in the LLC layer. Error correction is not possible, but errors can be detected. Two sub-modes are available in Unacknowledged mode: 2.1 Protected Mode FCS calculation across the header and user data 2.2 Unprotected Mode FCS calculation across the header and only a few of the first octets in the user data field.
ACTERNA
Page 7-6
Data Link Connection Identifier (DLCI) Data Link Connection Identifier (DLCI)
Mobile station SGSN
DLCI = SAPI + TLLI

SAPI z SAPI y
LLC
SAPI x
LLC
TLLI X
RLC/MAC BSSGP
TLLI Y
Page 7-7
The Data Link Connection Identifier (DLCI) consists of the parts SAPI and TLLI. The Service Access Point Identifier (SAPI) is used to differentiate between the various logical connections between the two LLC protocol instances in the SGSN and in the mobile station. The SAPI is transported in the header of every LLC frame. The Temporary Logical Link Identifier (TLLI) is used to address a particular mobile station. The TLLI is not a part of the LLC frame, but is part of the BSSGP messages (SGSN side) and of the RLC/MAC messages (MS side). Assignment of a TLLI to a particular MS is done by GPRS Mobility Management (GMM).
ACTERNA
Page 7-7
LLC Layer, Structure and SAPs LLC Layer, Structure and SAPs
GMM GMM
LLGMM LLGMM LL3 LL5
SNDCP SNDCP
LL9 LL11 TOM2
TOM TOM
TOM8
SMS SMS
LLSMS
LLC entity
LLC entity
LLC entity
LLC entity
LLC entity
LLC entity
LLC entity
LLC entity
LLC management
Multiplexer
LLC
BSSGP/RLC layer
Page 7-8
GRR
BSSGP
RLC/MAC (MS side) RLC/MAC (MS side)
BSSGP (SGSN side) BSSGP (SGSN side)

Data + signaling Signaling only
An LLC protocol instance consists of several so-called LLC entities. Each of these entities (sub-processes) carries out the LLC functions of Unacknowledged or acknowledged data transfer Flow control and Error monitoring or error correction independently of and without overlapping with other entities. As the SAPI, which represents an LLC or layer 2 address, is coded with 4 bits, it is possible in theory for up to 16 separate LLC connections to exist. A single mobile station can support several LLC connections simultaneously (e.g. SAPI 5 and SAPI 9). The multiplex procedure assigns the individual connections to the corresponding applications using the SAPIs and communicates the LLC frames with RLC/MAC or BSSGP. Insertion of the FCS and data ciphering are also handled by the multiplex procedure. LLC Management manages the resources needed for the individual LLC connections and processes errors.
ACTERNA
Page 7-8
LLC layer
SAPI:
11
Layer 3
LLC Frame Structure LLC Frame Structure

7 Address field (1 octet) Control field (variable length, max. 36 octets) PD 6 C/R 5 X 4 X 3 2 SAPI 1 0
Regardless of frame type this contains N(S), N(R), N(U), R(n) as well as other Control parameters
Information field (variable length, max. 1520 octets)
Contains the Layer 3 protocol data SNDCP, GMM, TOM and SMS
FCS (3 octets)
24-bit CRC
Page 7-9
Every LLC frame corresponds to the structure shown here. The address field contains the protocol discriminator (PD), the Command/Response bit, and the Service Access Point Identifier (SAPI). The protocol discriminator is fixed at 0; all other values are invalid. The C/R bit is coded as follows:
The control field is typically between 1 and 3 octets long. It can be up to 36 octets long in the case of a SACK frame. The control field mainly transports counters for flow control and detection of sequence errors. The PDUs of the higher protocol layers are encapsulated in the information field. The maximum length is 140 octets up to 1520 octets. The exact value may be negotiated. The Frame Check Sequence (FCS) is formed by a 24-bit wide CRC with the following generator polynomial: G(x) = x24 + x23 +x21 +x20 +x19 +x17 +x16 +x15 +x13 +x8 +x7 +x5 +x4 +x2 +1
ACTERNA
Page 7-9
Information Transfer Format (I Format) Information Transfer Format (I Format)

Address field
7 0
6 A
5 X
2 N(S)
Control field N(S) N(R) X N(R) S1 S2
Information field
FCS
Page 7-10
The I frame contains the send sequence number N(S) and the receive sequence number N(R) in the control field. These two counters are used to monitor the transmit and receive order and also serve to detect frames that are lost in the transmission path. They are modulo 512 counters. N(S) is incremented every time an I frame is sent. N(R) is used in the backward direction to confirm the number of I frames correctly received so far. If the A bit (Acknowledgement Request) is set to 1, the far end station can be forced to return an acknowledgement immediately after this I frame is received. If the A bit is not set, the window size (max. number of as yet unacknowledged I frames) is between 2 and 16 frames, depending on the SAPI. The S1 and S2 bits are Supervisory Function bits. They are used to distinguish between RR, RNR, ACK and SACK control frames. As an I frame can also be used in combination for acknowledgements because of the N(R) counter in the control field this is often referred to as I+S format. Acknowledgement of receiver I frames will only have to be made using RR frames in the event that there is no data to transmit in opposite direction at the moment. Otherwise, the current receive sequence number N(R) is sent in an I frame.
ACTERNA
Page 7-10
Supervisory Format (S Format) Supervisory Format (S Format)

Address field
7 1
6 0
5 A
4 X
3 X
1 N(R) S1
Control field N(R) S2
FCS
S1 0 0 1 1
S2 0 1 0 1
Frame type RR ACK RNR SACK
Page 7-11
The S frame is used to acknowledge the I frames so far correctly received in the opposite direction. The S frame is further differentiated by the S1 and S2 bits. Receive Ready (RR) Confirms correct receipt of N(R) I frames and indicates readiness to receive. Receive not Ready (RNR) Confirms correct receipt of N(R) I frames and indicates that no further frames can be received at the moment. Acknowledgement (ACK) Basically corresponds to RR/RNR but without signaling receive readiness. Selective Acknowledgement (SACK) This allows selective acknowledgement or request for repeat of individual frames out of a group of 255 I frames.
ACTERNA
Page 7-11
Selective Acknowledgement Frame (SACK Frame) Selective Acknowledgement Frame (SACK Frame)
7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 Address field 1 0 A X X N(R) 1 K
R(4) R(5) R(6) R(7) R(8)
Address field 0 1 A X X N(R) X SACK bitmap

R(1)
N(S) N(R) 1 K
R(4) R(5) R(6) R(7) R(8)
N(R) X
R(1)
N(S)
X
R(2)
X
R(3)
X
R(2)
X
R(3)
R(9) R(10) R(11) R(12) R(13) R(14) R(15) R(16)
.....
R(9) R(10) R(11) R(12) R(13) R(14) R(15) R(16)
R(249) R(250) R(251) R(252) R(253) R(254) R(255)
.....
R(249) R(250) R(251) R(252) R(253) R(254) R(255)
FCS
S frame SACK format
Information field
FCS
I frame SACK format

Page 7-12
Selective Acknowledgement can be contained in an I frame or only in an S frame without information field. The diagram shows the different control field formats in each case. The length of an I frame control field with SACK bitmap is 36 octets maximum. SACK bitmap: Every bit from R(1) to R(255) in the SACK bitmap corresponds to one of 255 received I frames. The basis for this group of I frames is specified with N(R). R(1) thus stands for frame N(R) +1 and N(255) stands for N(R) + 255. Setting a bit R(n) = 1 acknowledges correct receipt of the frame N(R)+n. R(n) = 0 means that the I frame R(n) + n was not received correctly and must be repeated. The length of the SACK bitmap is given by the length indicator K. The maximum length is 32 bytes. K value range = 0 to 31 (length of the SACK bitmap = K+1).
ACTERNA
Page 7-12
Unconfirmed Information Format (UI Format) Unconfirmed Information Format (UI Format)
Address field
7 1
6 1
5 0
4 X
3 X
1 N(U) E
Control field N(U) PM
Information field
FCS
Page 7-13
This frame type is used for exchanging unacknowledged and thus unprotected information in respect of the send sequence and any frame losses. Any frame losses or interference in the send or receive sequence bust be detected and corrected if necessary by the higher protocol layers. Transmitter unconfirmed sequence number N(U) Send sequence number for unacknowledged frames. Modulo 512 counter. Encryption function bit (E) If this bit is set to 1, the information field and FCS are transported in encrypted (ciphered) form. If the E bit = 0 ciphering does not take place in the transport path. Protected mode bit (PM) Indicates whether the FCS should be calculated from the complete LLC frame from the start of the address field to the end of the information field (PM = 1) or if the address field, control field and the first 4 octets of the information field are to be used in calculating the FCS (PM = 0).
ACTERNA
Page 7-13
Unnumbered Format (U Format) Unnumbered Format (U Format)
Address field
7 1
6 1
5 1
4 P/F
3 M4
2 M3
1 M2
0 M1
Control field
Information field (optional for SABM, UA and FRMR) M4 0 FCS 0 0 0 1 1 M3 1 1 0 1 0 0 M2 1 1 0 0 0 1 M1 1 0 1 0 0 1 Frame type SABM UA DM DISC FRMR XID
Page 7-14
SABM (Set Asynchronous Balanced Mode) . Used to set up an LLC connection for a particular SAPI. After receiving a SABM message, the opposite end also initializes the necessary variables, etc. and acknowledges SABM with a UA message. An optional information field contains XID parameters for negotiating LLC and /or layer 3 parameters. In this case the UA also contains XID parameters in the information field. UA (Unnumbered Acknowledge) Positive acknowledgement of SABM and DISC commands. DM (Disconnected Mode) Negative acknowledgement of a SABM command (the LLC connection cannot be set up in this case). DISC (Disconnect) Indicates the disconnection of an LLC connection. DISC is acknowledged with UA. FRMR (Frame Reject) Response to receipt of a (syntactically) incorrect frame. The frame error cannot be cleared by repeating the frame. The information field (maximum length 10 octets) contains information about the cause of the rejection. XID (Exchange Identification) Used to negotiate LLC and L3 parameters if this has not already been done with SABM/UA. The Poll/Final bit is used to request an immediate receive confirmation.
ACTERNA
Page 7-14
Exchange Identification Frame (XID) Exchange Identification Frame (XID)

7 6 5 4 3 2 1 0 Address field 7 Control field (1 byte) 1 6 1 5 1 4 P/F or XL XID-Parameter Type LI MS byte ... LS byte X LI 1 X 1 1 1 P/F 1 1 P/F or 0 1 1 0 UA 0 1 1 1 SABM 3 1 2 0 1 1 0 1 XID
Further XID parameters
FCS
Page 7-15
This frame type is used for negotiating the LLC and Layer 3 parameters between the two protocol instances concerned. The negotiation procedure takes place in one step. One side makes the other side an offer. To do this, it transports the parameter to be negotiated in the information field of an XID or SABM message. The other side can accept this parameter by returning an exact copy of the information field in an XID or UA message, or it can put other values into its response to the negotiating partner. These values must be accepted by the other end. All parameters that are not to be negotiated retain their previously set values or default values and are not transmitted in the information field. XID parameter format: XL = 1: Length indicator = 2 bits wide. The following octet already contains parameter data in this case. XL = 0: Type: LI: Length indicator = 8 bits wide. Parameter name. Length indicator for the subsequent octets.
ACTERNA
Page 7-15
Negotiable LLC parameters
ACTERNA
Page 7-16
Overview of Frame Types Overview of Frame Types

Format Command frame RR Information and control frames RNR SACK DISC Unnumbered frames SABM XID RNR SACK DM UA FRMR XID 1 1 0 1 0 0 0 0 1 1 0 1 1 1 0 0 0 0 1 1 0 1 1 0 0 1 0 1 ACK Response frame S1 RR ACK 0 0 S2 0 1 Coding M4 M3 M2 M1 -
Page 7-17
Notes:
ACTERNA
Page 7-17
TLLI Assignment TLLI Assignment

Mobile station Mobile station SGSN SGSN
GMM
GMM
LLGMM-ASSIGN-REQ
LLGMM-ASSIGN-REQ
LLC RR
(RLC/MAC)
LLC
BSSGP
Page 7-18
The assignment of TLLIs takes place in the same manner in the MS and in the SGSN. GPRS Mobility Management sends a LLGMM-ASSIGN-REQ service primitive with the value of the new TLLI to LLC Management. The LLC layer stores the assigned TLLI and communicates it between the LLC and the subordinate protocol layers (RR or BSSGP) if required. The TLLI is calculated directly from the P-TMSI. There is therefore a direct, unique, one-to-one relationship between the P-TMSI and the TLLI within a routing area. This procedure preserves the confidentiality of the subscriber identity.
ACTERNA
Page 7-18
ABM Establishment Procedure ABM Establishment Procedure

Transmitter Transmitter Receiver Receiver
L3
LLC
LLC
L3
LL-ESTABLISH-REQ SABM LL-ESTABLISH-IND LL-ESTABLISH-RES UA LL-ESTABLISH-CNF
Internal interface between protocol layers
ABM = Asynchronous Balanced Mode

Page 7-19
The establishment of a layer 2 connection (LLC connection) takes place by triggering a LLC user on the transmit side that wants to make use of acknowledged communications (ABM) from LLC. For this purpose, the corresponding primitives LL-ESTABLISH-REQUEST and LL-ESTABLISHINDICATION are transported between the layers. Confirmation is indicated by the primitives LL-ESTABLISH-RESPONSE and LL-ESTABLISH-CONFIRMATION. The control messages SABM and UA are exchanged on the communication interface.
ACTERNA
Page 7-19
Negotiating LLC Parameters Negotiating LLC Parameters

Transmitter Transmitter Receiver Receiver
L3
LLC
LLC
L3
XID
XID LL-XID-IND LL-XID-IND
Internal interface between protocol layers
Page 7-20
Negotiation of LLC parameters is possible in ABM (Asynchronous Balanced Mode) and ADM (Asynchronous Disconnected Mode). The parameters are negotiated by exchanging XID messages. The opposite side can accept the suggested values or replace them with its own values. The parameter values must remain within specified value ranges and be changed in a specific direction. As a rule, parameters are only changed if one peer cannot meet the requested requirements. The XID message contains all of the parameters to be negotiated. If the parameter changes affect the layer 3 protocol instances, these are informed by means of LL-XID-IND primitives.
ACTERNA
Page 7-20
LLC Parameters LLC Parameters

LLC version Input offset value (IOV) = random number for data ciphering
Retransmission timers (T200 and T201) Maximum number of repeat transmissions (N200) Maximum number of octets in an information field (N201) Maximum number of octets in a L3 header (N202) Maximum size of I frame buffer (m) Window size for frame acknowledgements (k)
Page 7-21
Default values of LLC layer parameters:
ACTERNA
Page 7-21
Notes:
ACTERNA
Page 7-22
GMM
(GPRS Mobility Management)
Page 7-23
Notes:
ACTERNA
Page 7-23
GMM Reference Model GMM Reference Model

GMM
SAP RR SAP GMMRR SAP LLGMM SAP LLGMM
GMM
SAP GMM
LLC
Relay
LLC BSSGP Network service L1
RR (RLC/MAC) GSM RF
RLC/MAC
BSSGP Network service L1
GSM RF
MS
Um
BSS
Gb
SGSN
Page 7-24
The diagram shows the architecture of GPRS Mobility Management as referred to the MS and SGSN. The BSSGP layer in the BSS also includes an interface to the GMM layer above it. This is not shown here to preserve clarity (see section on BSSGP). In the MS: GMM has direct access to the services of the RR sub-layer (which in turn includes the RLC and MAC entities). The RR and GMMRR service access points are defined for this. The services of the LLC layer are utilized via the LLGMM SAP. In the SGSN: LLC services can be accessed via the LLGMM SAP. A direct interface to BSSGP is available via the GMM SAP. GMM ensures that a particular MS is tracked as it moves through the network, cells are changed and the current location is always known to the network. GMM processes the GPRS attach / detach procedures.
ACTERNA
Page 7-24
SNDCP
(Sub-network Dependent Convergence Protocol)
Page 7-25
Notes:
ACTERNA
Page 7-25
SNDCP Protocol Stack SNDCP Protocol Stack
Higher layers (IP or X.25) SNDCP
Relay
SNDCP
GTP
LLC
LLC Lower layers
UDP / TCP IP Lower layers
Lower layers
MS
SGSN
Page 7-26
The Subnetwork Dependent Convergence Protocol (SNDCP) serves for transporting user data between the SGSN and MS and uses the services of the LLC layer to do so. Main functions of SNDCP: Multiplexing different PDPs, i.e. parallel transport of different data packets to and from different packet networks. PDP = Packet Data Protocol (e.g. X.25 or IP). Compression and decompression of user data. Compression and decompression of protocol control information. Segmenting the data packets before transmission Reassembling the data packets after reception
ACTERNA
Page 7-26
Multiplexing Different Protocols Multiplexing Different Protocols
Packet data protocol 1 (e.g. IP)
Packet data protocol 2 (e.g. IP)
Packet data protocol 3 (e.g. X.25)
N-PDU
NSAPI
SNDCP SNDCP
SAPI
SN-PDU
LLC LLC
Page 7-27
The protocol layers above SNDCP consist of generally applicable network protocols such as IP or X.25, etc. The individual applications or different network accesses are identified by means of the Network Layer Service Access Point Identifier (NSAPI). Multiplexing of the network protocols is possible on this basis. All network protocols use the services of a single, common SNDCP instance, so PDP multiplexing is unavoidable. Multiplex formation in the LLC layer is by means of the SAPIs. Several PDPs can share a logical LLC connection as shown in the example. A PDP can have more than one PDP context and hence several NSAPIs. Each assigned NSAPI makes use of LLC layer services via the SAPI. Several NSAPIs can access the LLC layer via a single, common SAPI. The data packets in the network layer are called Network Protocol Data Units (N-PDU). The data packets in the SNDCP layer are called SNDCP Protocol Data Units (SN-PDU).
ACTERNA
Page 7-27
PDP Multiplexing PDP Multiplexing
PDP type IP X.25 IP
Assigned NSAPI 12 13 14
PDP address 133.12.75.111 13254 133.12.48.213
Example of linking between NSAPI and PDP address
Page 7-28
Notes:
ACTERNA
Page 7-28
The SAPs Around SNDCP The SAPs Around SNDCP

SNDCP user
Control Control entity entity PDP PDP or or Relay Relay PDP PDP or or Relay Relay PDP PDP or or Relay Relay
...
15
NSAPI
SNDCP SNDCP layer layer
SNDCP management entity
SNDCP entity
SNSM
LL3
LL5
LL9
LL11
SAPI
Session Session management entity management entity
LLC layer LLC layer
Page 7-29
As shown in this diagram, only the PDP services are routed via the SNDCP layer. All other services, such as SMS or signaling, do not use any SNDCP services but address the layer 2 services directly using the corresponding service access point. NSAPI value range: 0 Escape mechanism for future applications 1 Point-to-multipoint transmission 2-4 Reserved for future applications 5-15 Dynamically assigned value The SAPI values 3, 5, 9, and 11 are reserved for the SNDCP layer for access to LLC, i.e. LLC provides SNDCP with a maximum of 4 simultaneous logical connections distinguished by SAPI.
ACTERNA
Page 7-29
SNDCP Model SNDCP Model

Network layer Network layer
H D N - PDU H D
Header compression
Header decompression
User data compression
SNDCP SNDCP
User data decompression
Segmentation
Reassembly
SN - PDUs
LLC LLC
H = Header, D = Data
Page 7-30
To be as economical as possible with available resources and bandwidths, SNDCP compresses the protocol information (header compression) and the subscribers user data before transporting it. If necessary, overlong user data packets are matched to the maximum size allowed on the LLC layer by segmenting them. Compression of protocol header and user data are optional SNDCP functions.
ACTERNA
Page 7-30
Segmentation Segmentation
N-PDU
Header
Data
H F=1 M=0
Data
H F=0 M=0
Data
H F=0 M=0
Data
H F=0 M=1
Data
Page 7-31
All the data packets transferred from the network layer to the SNDCP layer are buffered in the SNDCOP layer before being compressed and segmented. Network layer PDPs are segmented into several parts in the SNDCP layer if necessary. This is the case if the layer 3 data packets exceed the maximum size defined in N201 (LLC parameter). The default value of N201 depends on the type of transmission (Acknowledged mode or Unacknowledged mode) and on the SAPI and is between 270 and 1503 octets. All the segments thus formed include a More Data bit (M bit), which indicates the last segment to the receiver (M=1). The first segment is indicated by the F bit (=1). It is set to 0 in all other segments. When segmented packets are received, the individual parts, starting with F=1 and ending with M=1, are buffered in the SNDCP layer and reassembled.
ACTERNA
Page 7-31
Compression Compression
SNDCP user
PDP PDP or or Relay Relay PDP PDP or or Relay Relay PDP PDP or or Relay Relay PDP PDP or or Relay Relay
...
15
NSAPI
SNDCP layer SNDCP layer

Protocol compression Data compression Data compression Protocol compression Protocol compression
Data compression
LL3
LL5
LL9
LL11
SAPI
LLC layer LLC layer
Page 7-32
Several so-called compression entities that perform either protocol or data compression are available within SNDCP (the compression entities can be considered as part processes of a program). The compression entities correspond to a certain SAPI. The entity number is used to identify a compression entity. This is indicated to SNDCP by the SNDCP user within the XID parameter. Protocol compression compresses the header of a N-PDU. At present, GPRS specifications only describe compression of TCP/IP headers for IP applications and refer to the IP specification RFC1144. The procedures conforming to V.42bis are used for data compression. The compression parameters for a particular compression entity can be negotiated using the XID mechanism.
ACTERNA
Page 7-32
Negotiating L3 Parameters Negotiating L3 Parameters

Transmitter Transmitter SNDCP user Receiver Receiver
SNDCP
LLC
LLC
SNDCP
SNDCP user
SN-XID-REQ LL-XID-REQ XID LL-XID-IND SN-XID-IND SN-XID-RES LL-XID-RES XID LL-XID-CNF SN-XID-CNF
Page 7-33
XID parameters are used to match the SNDCP services to GPRS mobile communications. The individual parameter values, such as the compression method, are negotiated using special SNDCP exchange ID (XID) parameters, i.e. parameters are suggested by a (Request) message which the communications destination either reflects (= accepts) or alters and returns (see diagram above). This parameter negotiation can be initiated by the mobile terminal or by the SGSN and is intended to improve the efficiency of the connection, i.e. to increase the transmission rate. At present, only compression parameters are negotiated by the XID procedure.
ACTERNA
Page 7-33
The two following tables show the negotiable compression parameters
a) TCP/IP header compression parameters
b) Parameters for V.42bis data compression
P0: 00 = No compression 01 = Compression direction MS -> SGSN 10 = Compression direction SGSN -> MS 11 = Compression in both directions P1: Maximum number of code words in compressor P2: Maximum number of codable octets
ACTERNA
Page 7-34
Acknowledged Data Transfer Acknowledged Data Transfer

SNDCP
LLC
SABM UA
LLC
SNDCP
SNDCP user
LL-ESTAB-IND
LL-ESTAB-IND
SN-DATA-REQ LL-DATA-REQ Data transfer (I frame) LL-DATA-IND LL-DATA-CNF Acknowledgement (e.g. RR) SN-DATA-IND
Page 7-35
Transmission of GPRS packets in this mode is by means of a pseudoacknowledged service, since the SNDCP receiver does not acknowledge the packets. Acknowledgment is already made through layer 2 (LLC) of the receiver. As acknowledgement takes place in the LLC layer, this unacknowledged, at least from the point of view of the SNDCP user. service is
LLC saves the data packets, which are transported as I frames, until it receives an acknowledgement, e.g. in the form of a Receive Ready (RR) frame. SNDCP must store the N-PDUs from the network layer in its buffer until all the pertinent segments have been correctly transmitted and have been positively acknowledged by LLC.
ACTERNA
Page 7-35
Unacknowledged Data Transfer Unacknowledged Data Transfer

SNDCP
LLC
LLC
SNDCP
SNDCP user
SN-UNITDATAREQ LL-UNITDATAREQ Data transfer (UI frame) SN-UNITDATAIND LL-UNITDATAIND
Page 7-36
The unacknowledged mode shows the normal sequence of events: There are only REQUEST and INDICATION primitives through the individual layers. Once the data packets have been transferred to the next lower protocol layer, they are deleted from the SNDCP and LLC buffers. No repetition of packets takes place in the event of an error. This function must be provided by higher protocol layers if it is required.
ACTERNA
Page 7-36
SN-PDU Format SN-PDU Format

7 1 byte 1 byte 1 byte X 6 F 5 T 4 M 3 2 1 0
NSAPI PCOMP 7 6 F 5 T 4 M 3 2 1 0
DCOMP N-PDU number
1 byte n bytes Data segment 1 byte 1 byte Format for Acknowledged Mode (SN-DATA PDU format) 1 byte
NSAPI PCOMP N-PDU number
DCOMP Segment number
N-PDU number
n bytes
Data segment
Format for Unacknowledged Mode (SN-UNITDATA PDU format)
Page 7-37
More bit (M) 0 further segments of a N-PDU follow 1 last segment of a N-PDU SN-PDU type (T) 0 SN-DATA-PDU 1 SN-UNITDATA-PDU First segment indicator (F) 0 not the first segment of a N-PDU 1 first segment of a N-PDU NSAPI 0 to 15 Data compression coding (DCOMP) 0 no compression 1-14 pointer to a negotiated type of compression Protocol control information compression coding (PCOMP) 0 no compression 1-14 pointer to a negotiated type of compression Segment number 0-255 sequence number for the segments of a N-PDU N-PDU number 0-255 sequence number of the N-PDUs in acknowledged mode 0-4095 sequence number of the N-PDUs in acknowledged mode
ACTERNA
Page 7-37
The following table shows a list of all service primitives and their parameters for layer communication with the SNDCP entity.
ACTERNA
Page 7-38
TOM
(Tunneling of Messages)
Page 7-39
Notes:
ACTERNA
Page 7-39
TOM Envelope Format TOM Envelope Format
Protocol discriminator (1) Protocol header max. 14 bytes
LI
Message envelope max. 220 bytes
Page 7-40
Tunneling of Messages (TOM) is an optional general protocol layer for exchanging transparent, application-specific signaling messages between the network and the MS. TOM uses 2 LLC SAPs, one for high priority and one for low priority messages. The basic format of a TOM message consists of a variable-length protocol header (max. 14 bytes) followed by a variable-length message capsule (max. 220 bytes). The LI indicates the length of the protocol header. The protocol discriminator has the value 1, all other values are invalid or reserved for future applications. The protocol header contains information about the specific application that is using TOM. The message capsule contains the payload, i.e. the transparent message from or to the MS. TOM is a versatile protocol, e.g. it is possible to exchange messages between an MS and a non GSM MSC via the GSM network components.
ACTERNA
Page 7-40
SMS
(Short Message Service)
Page 7-41
Notes:
ACTERNA
Page 7-41
GPRS SMS Transfer GPRS SMS Transfer
SGSN SGSN
HLR
SMS SMS gateway MSC gateway MSC
SMS service center
SMS Get routing info Routing info
SMS SMS
SMS result SMS result Delivery report
Page 7-42
SMS accesses LLC via SAPI 7. The SMS protocol for GPRS corresponds to the definitions already familiar from GSM. A GPRS attached MS receives the SMS messages via the GPRS radio channels. If the MS is both IMSI and GPRS attached, the SMS can be transmitted via GPRS channels or via the GSM signaling channels (SDCCH).
ACTERNA
Page 7-42
Chapter 8 Chapter 8
The Tunnel Principle and the GPRS Tunneling Protocol
Theres a tunnel ahead Its dark inside but on the outside its light, instead
(Anonymous)
Page 8-1
Notes:
ACTERNA
Tunnel Principle and GTP
Page 8-1
Protocol Stack Protocol Stack

Application IP/X.25 Relay SNDCP SNDCP LLC RLC RLC MAC GSM RF MAC GSM RF Relay BSSGP BSSGP Network service L1bis Network service L1bis IP L2 L1 IP L2 L1 LLC GTP UDP/ TCP UDP/ TCP GTP IP/X.25
MS
Um
BSS
Gb
SGSN
Gn
GGSN
Gi
Page 8-2
At the Gn interface, all the protocol layers below the GTP layer are realized by an IP-based network. This IP network is called an intra-PLMN backbone network. The physical layer can be implemented as Ethernet, FastEthernet, ATM etc., as with any other IP network. The intra-PLMN backbone network is a private IP network that is exclusively reserved for transmitting GPRS data packets and GPRS signaling packets. There is no connection to the public data network (e.g. Internet). All network nodes (SGSN, GGSN) in a single PLMN are connected together by the intra-PLMN backbone network. Network node addressing is based on an own local IP address area. This address area must not be confused with the dynamic (or fixed) IP addresses that are assigned to the mobile station. The GPRS Tunneling Protocol (GTP) serves to transfer the application data between the MS and GGSN over the Gn segment. For this, the application packets (IP or X.25) are encapsulated in GTP frames and passed on in this form to the subordinate TCP layer. To be precise, IP packets are encapsulated in GTP packets which are in turn encapsulated in IP packets for transmission on the backbone network. This is known as a transparent tunnel mechanism.
ACTERNA
Page 8-2
The Tunnel Principle The Tunnel Principle

Preface
IP application
IP application
GPRS network Tunnel Gateway Gateway IP network
Page 8-3
The tunnel principle is a method of transporting data from one point to another by using a protocol to encapsulate the data of another protocol. This method is used by many protocols; for example the ISUP encapsulates various DSS1 messages in DSS1 format such as HLC or LLC parameters into the ISUP Access Transport parameter. The transmitters and receivers of encapsulated messages generally have no inkling of the other protocol that they are transmitting. GPRS uses the tunnel principle to transport any user data over the GPRS network, i.e. X.25 or IP packet types or other formats are transported over the GPRS network to the destination network (e.g. IP or X.25). An IP application is attached to an IP fixed network (right-hand side). The mobile and fixed network applications communicate using IP. Since the GPRS network does not, however, support this protocol, IP frames must tunnel through the GPRS network transparently (encapsulated). The advantage of this principle is that the GPRS network is independent of the various transmission protocols used by different networks. GPRS can thus transport any possible (including future)type of packet format without problems.
ACTERNA
Page 8-3
GPRS Tunnel Principle (1) GPRS Tunnel Principle (1)

... shut your eyes and go through
Radio part SNDCP tunnel

(radio subnet)
GPRS part
PSPDN
SGSN SGSN GTP tunnel GGSN GGSN
NSAPI + TLLI
TID (= IMSI + NSAPI)
Page 8-4
In fixed networks, the tunnel principle is bound to a transport mechanism, I.e. the transporting protocol. The situation is somewhat more complicated in the mobile network: First of all, a radio tunnel is required on the radio interface, followed by the GPRS tunnel. The radio tunnel is also called the SNDCP tunnel, since the adjustment functions of the Subnetwork Dependent Convergence Protocol are used here and the SNDCP layer provides the tunnel. An SNDCP tunnel that connects a particular mobile station to the network is addressed by the TLLI and NSAPI. On the fixed network side of the PLMN, the GPRS tunneling protocol generates the GTP tunnel. The GTP tunnel is addressed by its Tunnel Identifier (TID), which consists of the IMSI and NSAPI.
ACTERNA
Page 8-4
GPRS Tunnel Principle (2) GPRS Tunnel Principle (2)

Tunnel access to different networks
GPRS Radio part SNDCP tunnel

(radio subnet)
GTP tunnel SGSN SGSN GTP tunnel GTP tunnel part
X.25 Internet Intranet
Page 8-5
The GPRS network can consist of several gateways to the same or different kinds of data network, i.e. the GPRS user can make connections to all these networks via the SNDCP tunnel. However, a GPRS user can only access a particular data network through the corresponding Gateway Support Node. The GGSNs that are possible are stored with the subscriber profiles in the HLR, with one GGSN address for every network. A GPRS mobile subscriber can only exchange data with different networks via an SNDCP tunnel. The combination of SNDCP tunnel and GTP tunnel is also called the Packet Data Protocol (PDP) context. Firstly, a radio connection must be established so that the desired PDP context can be set up, after which the data can be exchanged with a data network. Here, a terminal can serve several PDP contexts simultaneously. The transmitted and received data packets are called Packet Data Units (PDUs). Currently, two different types of PDU are defined: X.25 packets for the X.25 network and IP packets for the Internet. The tunnel principle makes transport of further packet types for future networks possible.
ACTERNA
Page 8-5
GPRS Tunneling Protocol GPRS Tunneling Protocol
IMSI TLLI PDP address TID SGSN and GGSN address RAI CI
User data + signaling
Page 8-6
Communication between the SGSN and GGSN (Gn interface) is based on the Internet protocol, with a GTP (GPRS Tunneling Protocol) application superimposed on it. This application enables transport and encapsulation of PDP (Packet Data Protocol) packets through the GPRS network elements up to the corresponding applications in the fixed network and the mobile terminal. GTP provides two layers for this: A signaling layer, which provides all the management functions for a PDP context, and a transport layer, which enables transport of the user data through the GPRS network nodes.
ACTERNA
Page 8-6
GPRS Tunneling Protocol The data packets are routed via tunnel pairs (SNDCP - GTP). GPRS uses socalled contexts in the network nodes (SGSN and GGSN) for selecting the correct tunnel. The addresses that play a part in this routing process (and hence in the context of this connection) are: IMSI (International Mobile Subscriber Identity): This parameter is described in detail in GSM 03.03. Every mobile terminal has a unique IMSI. TLLI (Temporary Logical Link Identity): This layer 2 address uniquely identifies the service access point (SAP), e.g. IP address SAP or X.25 address SAP. PDP (Packet Data Protocol) address: A GPRS subscriber (identified by IMSI in the GSM layer) can activate one or more layer 3 addresses. These can be assigned temporarily or permanently. Activation and deactivation takes place via MM (Mobility Management) procedures. TID (Tunnel Identifier): The GTP identifies a Packet Protocol Data context with this address. A TID consists of IMSI and NSAPI (Network Service Access Point Identifier). SGSN and GGSN addresses: Both GPRS network elements are identified by IP addresses. IPv4 or IPv6 address formats are used. RAI (Routing Area Identity): One or more cells are addressed using this address. This allows a mobile terminal to detect when it crosses a routing area frontier and a routing area update becomes necessary. RAI is transmitted as a broadcast message. CI (Cell Identifier): Defines a radio cell.
ACTERNA
Page 8-7
GTP header GTP header

Octet 1 2 3-4 5-6 7-8 9 10 11 12 13-20 7 6 Version 5 4 PT Message type Length SN Flow label SNDCP N-PDU number Spare Spare Spare TID Further information elements or T-PDU for transmitting user data 3 2 Spare 1 0 SNN Bit
Page 8-8
Version indicates the protocol version PT (Protocol Type): As well as GTP, there is also GTP. This bit indicates which type of GTP is involved (GTP = 1, GTP = 0). The SNN flag indicates whether or not the GTP header contains a SNDCP number. Message Type designates the corresponding GTP message. Length indicates the length in bytes of the GTP message after the header (GPDU). The SN (Sequence Number) is handled differently depending on whether GTP signaling messages or user data is being transported. For user data: SN = consecutive numbering For signaling: SN = identifies a signaling transaction The Flow Label serves to uniquely identify a GTP data flow The SNDCP N-PDU number is needed during an inter-SGSN routing area update procedure to coordinate the data traffic between the SGSNs affected and the MS. TID is the Tunnel Identifier, which refers to an MM and a PDP context.
ACTERNA
Page 8-8
GTP Messages GTP Messages

Path management messages
Tunnel management messages
GTP
Location management messages
GTP
Mobility management messages
T-PDU
GSN GSN
GSN GSN
Gn interface
Page 8-9
The GTP message T-PDU is available for transmitting the user data (Transmission Protocol Data Units, T-PDUs). All the remaining GTP messages can be divided into Path Management, Tunnel Management, Location Management and Mobility Management areas. Path Management has, among other things, the task of continuously checking the availability of the GTP connection using appropriate mechanisms (Echo Request and Echo Response messages). Tunnel Management takes care of setting up, updating and deleting the PDP contexts. Messages such as Create PDP Context Request, Delete PDP Context Request etc. have been defined for processing the communications necessary for this. If a GGSN is not connected to the HLR via a Gc interface (which is not strictly necessary in a GPRS network), GGSN HLR communications must also take place via the GTP. The GTP Location Management messages were defined for this purpose. GGSN HLR communications would be SS#7 / MAP-based via Gc. The GTP messages for Mobility Management transport the GSM 03.60 and GSM 04.08 MM messages between the GSNs. This includes all the messages for the GPRS attach procedure and for the inter-SGSN routing area update procedure, for example.
ACTERNA
Page 8-9
Overview of GTP Signaling Messages
The Message T-PDU is used to transmit user data. All other messages carry GTP signaling information.
ACTERNA
Page 8-10
Routing and Context Parameters Routing and Context Parameters

What role do context parameters play in routing?
SGSN SGSN GGSN GGSN
SNDCP-PDU {TLLI, NSAPI, PDP-PDU}
GTP-PDU {TID, PDP-PDU} PDP-PDU {PDPaddress}
SNDCP
GTP
GTP-PDU {TID, PDP-PDU}
IP
External IP network
SNDCP-PDU {TLLI, NSAPI, PDP-PDU}
PDP-PDU {PDP address}
Page 8-11
The diagram shows a simple example of how data packets are routed over the SNDCP and GTP tunnel, and the role played in this by the individual context parameters. To allow the sequence, the individual contexts have to be set up first. GTP provides a range of messages for this, which can be used to set up and release a new context and to alter an existing context. The TID (Tunnel Identifier) points to a context. This context contains all necessary routing parameters, such as TLLI, NSAPI, etc.
ACTERNA
Page 8-11
Context Activation Context Activation
SGSN SGSN
GGSN GGSN
Activate PDP context request Security functions Create PDP context request Create PDP context response Activate PDP context accept
Page 8-12
In this example, the context activation procedure is initiated by the mobile terminal with the Activate PDP Context Request message. This message contains the following parameters: = Layer 3 service access point identifier NSAPI = IMSI TI = Packet protocol type PDP Type = e.g. Internet address PDP Address = Name of access point (network identifier) Access Point Name = Quality of service parameter profile QoS requested = Additional PDP parameters PDP Configuration Options The GGSN responds with a Create PDP Context Response message. Receipt of this message by the SGSN triggers an Activate PDP Context Accept message containing the following parameters: = IMSI TI = Packet protocol type PDP Type = e.g. Internet address PDP Address = Name of access point (network identifier) Access Point Name = Negotiated quality of service parameter profile QoS negotiated = Based on QoS negotiated Radio Priority = Additional PDP parameters PDP Configuration Options The SGSN can now transfer PDP packets between the mobile terminal and the GGSN and charges start to be registered.
ACTERNA Tunnel Principle and GTP Page 8-12
Context Deactivation Context Deactivation

... Thats all, folks!
SGSN SGSN
GGSN GGSN
Delete PDP context request
Delete PDP context response Deactivate PDP context request
Deactivate PDP context accept
Page 8-13
The existing context is deactivated to release the connection. This is done using the procedure shown in this example. Deactivation is initiated by the SGSN. The GGSN or the MS can, of course, also demand deactivation of a PDP context. The message arrow directions will change accordingly.
ACTERNA
Page 8-13
Glossar
Abbreviation 2MS 2MU 3G a/b AA AAL AAL5 AB ABM ABR AC ACCH ACK ACM ACS ACSE ACU ADM ADPCM ADSL AE AEF AFC AGC AGCH AMI AMPS AMR AN ANF ANM ANS ANSI AOC APDU APN APON ARFCN ARQ ASE ASE ASN.1 ASS ATM AUT AUTN B channel Meaning 2 Mbit/s leased line, structured 2 Mbit/s leased line, unstructured 3 rd Generation Analog 2-wire interface Anonymous Access ATM Adaptation Layer ATM Adaptation Layer type 5 Access Burst Asynchronous Balanced Mode available bit rate Authentification Center Associated Control Channel ACKnowledgement Address Complete Message Adjacent Channel/Carrier Suppression Association Control Service Element Antenna Combinary Unit Asynchronous Disconnected Mode adaptive delta pulse code modulation asymmetric digital subscriber loop Application Entity Additional Elementary Functions Automatic Frequency Control Automatic Gain Control Access Grant Channel alternative mark inversion advanced mobile phone service Adaptive Multirate speech codec Access Network Additional Network Feature Answer Message Answer American National Standards Institute Advice of Charge Application Protocol Data Unit Access Point Name ATM PON Absolute RF Channel Number Automatic Request for Retransmission amplified spontaneous emission Application Service Element Abstract Syntax Notation One Autonomous Signaling Sequence Asynchronous Transfer Mode Authentification Authentication Token 64 kbit/s bearer channel
Copyright ACTERNA
Page 1
H. Visel
Glossar
Abbreviation BB BCC BCC BCCH BCD BCS BEC BECN BER BG BH BIB B-ISDN BLA BLO Bm BN BNHO BOC BRA BS BS BS BS-AG-BLKS-RES BS-BCCH-SDCCH-COMB BSC BS-CC-CHANS BSCU BS-FREQ-NCELL BSI BSIC BSIC-NCELL BSN BS-PA-MFRMS BSS BSSAP BSSAP+ BSSGP BSSOMAP BTS BTSM BVC BVCI C C/R CA CA-CN Meaning Backbone Bearer Base Station Color Code bearer channel connection Broadcast Control Channel Binary Coded Decimal Block Check Sequence Backward Error Correction Backward Explicit Congestion Notification Bit Error Rate Border Gateway Block Header Backward Indicator Bit Broadband-ISDN Blocking Acknowledgment Blocking Traffic channel for full rate voice coder Bit Number Barring all outgoing calls except those to Home PLMN Bell Operating Company Basic Rate Access Base Station Bearer Service Billing System The number of blocks on each common control channel Combination of dedicated and associated control Base Station Controller The number of basic physical channels supporting Base Station Controller Unit Frequency of the RF carrier on which the BCCH of an Base Station Interface Base Station Identity Code BSIC of an adjacent cell Backward Sequence Number The number of multiframes between two transmissions of Base Station Subsystem BSS Application Part Base Station System Application Part + Base Station System GPRS Protocol BSS Operation & Maintenance Application Part Base Transceiver Station BTS Management BSSGP Virtual Connection BSSGP Virtual Connection Identifier Conditional command/response Cell Allocation Cell Allocation RF Channel Number
Copyright ACTERNA
Page 2
H. Visel
Glossar
Abbreviation CAI CA-NO CAP CAPI CAS CATV CB CBA CBD CBR CC CC CCAF CCBS CCCH GROUP CCF CCNR CCPE CCR CCS CCU CD CDIVC CDIVI CDM CDMA CDR CELL-BAR-ACCESS CELL-RESELECT-HYSTERESIS CFB CFCCH CFE CFID CFN CFNR CFU CG CGB CGBA CGF CGI CGU CGUA CHAP CI CI CI Meaning common air interface RF Channel Number of the BCCH in a Particular Cell carrierless amplitude/phase Common Application Program Interface Channel Associated Signaling cable television Call Barring Change Back Acknowledgment Change Back Declaration constant bit rate Country Code Call Control Call Control Agent Function Completion of Calls to Busy Subscribers Group of MSs in idle mode Call Control Function Completion of Calls on No Reply Control Channel Protocol Entity Continuity Check Request Common Channel Signaling Channel Codec Unit. Call Deflection Controlled Diversion Consult Controlled Diversion Immediate code-division multiplexing code division multiple access Call Detail Record Cell Access Barred RXLEV Hysteresis required for Cell Reselection Call Forwarding on Busy Compact Frequency Correction Channel Control Function Element Control Function ID Confusion Call Forwarding on No Reply Call Forwarding Unconditional Charging Gateway Circuit Group Blocking Circuit Group Blocking Acknowledgement Charging Gateway Functionality Cell Global Identification Circuit Group Unblocking Circuit Group Unblocking Acknowledgement Challenge Handshake Authentication Protocol Cell Identity Common Information Connection Incomplete
Copyright ACTERNA
Page 3
H. Visel
Glossar
Abbreviation CIC CID CK CK CLIP CLIR CLLM CLNP CLNS CM CMD CMIP CN CN CNF CNIP CNIR CO COA COLP COM CON CONF CONI CONN CONN ACK CONP CONS COO COT CPAGCH CPBCCH CPCCCH CPE CPG CPK CPNCH CPPCH CPRACH CPU CR CR CR CRC CRE CS CS Meaning Circuit Identification Code Circuit Identification Check Byte Cipher Key Calling Line Identification Presentation Calling Line Identification Restriction Consolidated Link Layer Management Connectionless Network Protocol Connectionless Network Service Connection Management Command Common Management Information Protocol Corporate Network Core Network Confirm Calling Name Identification Presentation Calling Name Identification Restriction Call Offer Change Over Acknowledge Connected Line Identification Presentation Complete Connect Add-On Conference Connect Number Identity Connect Connection Acknowledge Connected Name Identification Presentation Connection-oriented network service Change Over Order Continuity Compact Packet Access Grant Channel Compact Packet Broadcast Control Channel Compact Packet Common Control Channel customer premises equipment Call Progress Call Park Compact Packet Notification Channel (for PTM-M on CPCCCH) Compact Packet Paging Channel Compact Packet Random Access Channel Call Pick-Up Cadenced Ringing Call Reference Connection Request Cyclic Redundancy Check Call RE-establishment procedure Coding Scheme Circuit Switched
Copyright ACTERNA
Page 4
H. Visel
Glossar
Abbreviation CSA CSCH CSPDN CT CT1 CT2 CTP CTS CU CU CUG CW D channel DAR DATEX-J DATEX-M DATEX-P DB DCCH DCF DCN DCOMP DCS1800 DDD DDI DDM DE DECT DET DF DFE DHCP DISC DIST DL DL DLCI DLD Dm DM DMT DND DNDO DNIC DNS DP DPC Meaning carrier serving area Compact Synchronization Channel Circuit Switched Public Data Networks Call Transfer first-generation cordless telephony second-generation cordless telephony connection termination point Cordless Telephony System Cell Update Central Unit of a MS Closed User Group Call Waiting Signaling channel (data channel) Disable Autonomous Response Btx service in Germany Broadband service in Germany Packet service in Germany Dummy Burst Dedicated Control Channel Data Communication Function Data Communication Network Identifier of the user data compression algorithm Digital Communications System at 1800 MHz directional division duplexing Direct Dialing In directional division multiplexing Discard Eligibility digital European cordless telecommunications Detach Don't Fragment decision feedback equalization Dynamic Host Configuration Protocol DISConnect Call Distribution Data Link (layer) Downlink Data Link Connection Identifier Data Link Discriminator Control Channel Disconnected Mode discrete multitone Do Not Disturb Do Not Disturb Override Data Network Identification Code Domain Name System Dialed Pulse Destination Point Code
Copyright ACTERNA
Page 5
H. Visel
Glossar
Abbreviation DPNSS DQDB DRX DS DS DSE DSE DSL DSP DSS DTAP DTE DTMF DTX DTX DUP EA EA EAR EC Ec/No ECA ECMA ECO EDFA EGPRS EIR ERR ESP ETSI EX F F F/W FA FAA FAC FACCH FACCH/F FACCH/H FAR FB FBI FCS FDD FDDI FDM Meaning Digital Private Network Signaling System Distributed Queue Dual Bus Discontinuous Reception Mechanisms Digit Signal digital section Dialogue Service Element Data Switch Exchange digital subscriber loop digital signal processing Digital Subscriber System Direct Transfer Application Part Data Terminal Equipment Dual Tone Multi-Frequency (signaling) Discontinuous Transmission Discontinuous Transmission Mechanism Data User Part extension address External Alarms Enable Autonomous Response European Community Ratio of energy per modulating bit to the noise Emergency Change Over Acknowledgement European Computer Manufacturers Association Emergency Change Over erbium-doped fiber amplifier Enhanced GPRS Equipment Identity Register ERRor Encapsulating Security Payload European Telecommunication Standardization Institute Exchange Flag First segment indicator bit Firewall Full Allocation Facility Accepted Facility Fast ACCH Full rate Fast Associated Control Channel Half rate Fast Associated Control Channel Facility Request Frequency correction Burst Final Block Indicator Frame Check Sequence Frequency Division Duplex fiber-optic data distribution interface frequency-division multiplexing
Copyright ACTERNA
Page 6
H. Visel
Glossar
Abbreviation FDMA FEC FECN FER FEXT FFS FH FIB FIR FISU FM FN FN-MAX FOT FPM FR FR FRF FRJ FRMR FS FSN FTP FTS FTTA FTTB FTTC FTTCab FTTH F-Type FWM Gb G-CDR GEA GEA2 GFT GGSN GGSN Gi GLR GMM GMM/SM GMSC Gn Gp GPRS GPRS-CSI Meaning frequency division multiple access Forward Error Correction Forward Explicit Congestion Notification Frame Erasure Ratio far-end crosstalk For Further Study Frame Header Forward Indicator Bit finite impulse response Fill In Signal Unit frequency modulation Frame Number Maxima TDMA Frame Number Forward Transfer four-photon mixing frame relay Frame Relay Frame Relay Forum Facility Rejected FRaMe Reject Further Study Forward Sequence Number File Transfer Protocol feeder transmission system fiber to the apartment fiber to the business fiber to the curb fiber to the cabinet fiber to the home Frame Data Type four-wave mixing Interface between an SGSN and a BSC. Gateway GPRS Support Node Call Detail Record GPRS Encryption Algorithm GSM GPRS Algorithm 2 Generic Functional Transport Gateway GPRS Support Node Gateway GPRS Support Node Reference point between GPRS and an external PDN. Gateway Location Register GPRS Mobility Management GPRS Mobility Management and Session Management Gateway Mobile Switching Center Interface between two GSNs within the same PLMN. Interface between two GSNs in different PLMNs General Packet Radio Service GPRS CAMEL Subscription Information
Copyright ACTERNA
Page 7
H. Visel
Glossar
Abbreviation GPRS-SSF GRA GRR GRS Gs GSA GSM GSM GSM PLMN GSM-SCF GSN GTP GTP-C GTP-U HANDO HCS HDLC HDSL HDTV HDWDM HE HFC HLR HMSC HOLD HO-MARGIN HPLMN HPU HSCSD HSN I IAM IANA IC ICB ICMP ID IDN IDR IE IEC IETF IFID IHOSS IHOSS:OSP IIR IK Meaning GPRS Service Switching Function Circuit Group Reset Acknowledgement GPRS Radio Resources service access point Circuit Group Reset Interface between an SGSN and MSC. GSM System Area Global System for Mobile Communication Groupe Speciale Mobile GSM Public Land Mobile Network GSM Service Control Function GPRS Support Node GPRS Tunnelling Protocol GTP Control GTP User Handover Header Check Sequence High Level Data Link Control high-speed digital subscriber loop high-definition television high-density wavelength-division multiplexing Host Exchange hybrid fiber/coax Home Location Register Home Mobile-services Switching Center Call Hold SDL Message name for Handover Margin Home PLMN Hand Portable Unit High Speed Circuit Switched Data Hop Sequence Number Information (frames) Initial Address Message Internet Assigned Number Authority Interexchange Carrier Incoming Calls Barred Internet Control Message Protocol Identification Integrated Digital Network Identification Request Information Element International Electrotechnical Commission Internet Engineering Task Force Interface ID Internet-Hosted Octet Stream Service Internet Hosted Octet Stream Service:Octect Stream Protocol infinite impulse response Integrity Key
Copyright ACTERNA
Page 8
H. Visel
Glossar
Abbreviation IM IM_GSN IMEI IMSI INAP IND INTR INU IOV IP IP IPCI IPNS IPv4 IPv6 IPX IR IRS ISDN ISI ISO ISP ISPC ISUP ITU ITU-T IWF K Kc Ki Kl Ks KSI L2M L2TP LAC LAC LAI LAN LAPB LAPD LATA LCN LCS LE LED LI Meaning In-Call Modification Intermediate GSN International Mobile station Equipment Identity International Mobile Subscriber Identity Intelligent Network Application Part Indication Intrusion Interworking Unit Input Offset Value Intelligent Peripheral Internet Protocol ISDN-Port-Channel-Identification ISDN PBX Network Specification Internet Protocol version 4 Internet Protocol version 6 Internet Packet eXchange Incremental Redundancy Identification Response Integrated Services Digital Network intersymbol interference International Standardization Institute Internet Service Provider International Signaling Point Code ISDN User Part International Telecommunications Union International Telecommunication Union - Telephone Inter Working Function Constraint Length of the Convolutional Code Ciphering Key Individual subscriber authentication key Location Key session Key Key Set Identifier Layer 2 Management Layer-2 Tunnelling Protocol Location Area Code L2TP Access Concentrator Location Area Identity Local Area network Link Access Protocol Balanced Link Access Procedure on the D-channel Local Access and Transport Area Local Communication Network Location Services Local Exchange light-emitting diode Length Indicator
Copyright ACTERNA
Page 9
H. Visel
Glossar
Abbreviation LI LL LLC LLDN LLE LLGMM LLM LLME LL-PDU LL-PDU Lm LMDS LMWS LNS LOP LR LR LS LSP LSSU LT M M MA MAC MACN MAF MAIO MAN MAP MCC M-CDR MCI MCID MCS MD MDL ME MEF MIC MIP MLPP MM MME MMI MNC MNRF Meaning Line Information Logical Link Logical Link Control local line distribution network Logical Link Entity LLC to GPRS Mobility Management service access point Logical Link Management Logical Link Management Entity LLC PDU LLC PDU Traffic channel with capacity lower than Bm local multipoint distribution service licensed millimeter-wave service L2TP Network Server Loop prevention Local Reference Location Register Linkset Link Selector Parameter Link Status Signal Unit Line Termination Mandatory More bit used to indicate the last segment of N-PDU Mobile Allocation Medium Access Control Mobile Allocation Channel Number Mobile Additional Function Mobile Allocation Index Offset Metropolitan Area Network Mobile Application Part Mobile Country Code Mobility Management - Call Detail Record Malicious Call Identification Malicious Call Identification Modulation and Coding Scheme Mediation Device (mobile) Management (entity) - Data Link (layer) Maintenance Entity Maintenance Entity Function Mobile Interface Controller Mobile IP Multi-Level Precedence and Preemption Mobility Management Mobile Management Entity Man Machine Interface Mobile Network Code Mobile station Not Reachable Flag
Copyright ACTERNA
Page 10
H. Visel
Glossar
Abbreviation MNRG MNRR MNS MOC MPEG MPH MS MS ISDN MSC MSCU MSM MSN MS-RANGE-MAX MSRN MS-RXLEV-L MS-TXPWR-CONF MS-TXPWR-MAX-CCH MS-TXPWR-REQUEST MSU MT MT MTC MTP MTP2 MTP3 MTU MUMS MUX N/W NB NBIN NC NCC NCELL NCH NDC NE NE2SYNC NEF NEXT NF NGAF NI NM NMC NMG NMN Meaning Mobile station Not Reachable for GPRS flag Mobile station Not Reachable Reason Mobile Network Signalling Mobile Originating Call Motion Pictures Experts Group (mobile) Management (entity) - PHysical (layer) Mobile Station Mobile Station ISDN Number Mobile Switching Center Mobile Station Control Unit Multi-Slot-Map Multiple Subscriber Number Mobile Station Range Maximum Mobile Station Roaming Number Lower Receive Level MS Transmitted RF Power for Confirmation Maximum Allowed Transmitted RF Power for MSs MS Transmitted RF Power Request. Message Signal Unit Message Type Mobile Terminal Mobile Terminated Call Message Transfer Part Message Transfer Part layer 2 Message Transfer Part layer 3 Maximum Transmission Unit Multi User Mobile Station multiplexer Network Normal Burst A parameter in the hopping sequence Network Code Network Color Code Neighboring (adjacent) Cell Notification Channel (for PTM-M on CCCH) National Destination Code Network Element Network clock supply unit Network Element Function near-end crosstalk Network Function Non-GPRS Alert Flag Network Interception Network Management Network Management Center Network Management Gateway Network Management Node
Copyright ACTERNA
Page 11
H. Visel
Glossar
Abbreviation NMSI N-PDU NRM NS NS NS NSAP NSAPI NSC NSDU NSE NSEI NSS NS-SAP NS-VC NS-VCI NS-VL NS-VLI NT NTBA NTPM O O&M OACSU OCB OD ODN OFDD OFDM OFSK OLT OMAP OMC OMUP ONU OPC OS OSF OSI OSI RM OSP OSP:IHOSS OTN P PABX PACCH PAD Meaning National Mobile Station Identification number Network Protocol Data Unit Network Resource Management Night Service Network Service Network Service Network Service Access Point Network Layer Service Access Point Identifier Network Service Center Network Service Data Unit Network Service Entity Network Service Entity Identifier Network and Switching Subsystem Network Service Service Access Point Network Service Virtual Connection Network Service Virtual Connection Identifier Network Service Virtual Link Network Service Virtual Link Identifier Network Termination Network termination Basic rate Access Network Termination Primary Rate Access Optional Operations & Maintenance Off-Air-Call-Set-Up Outgoing Calls Barred Optional for operators to implement for their aim optical distribution network optical frequency-division duplexing optical frequency-division multiplexing optical frequency-shift keying optical line termination Operations Maintenance and Administration Part Operations and Maintenance Centre Operating and Maintenance User Part optical network unit Origination Point Code Operating System Operations System Function Open Systems Interconnection OSI Reference Model Octet Stream Protocol Octet Stream Protocol for Internet Hosted Octet Stream Service optical transition node Propose bit private automatic branch exchange Packet Associate Control Channel Packet Assembly/Disassembly facility
Copyright ACTERNA
Page 12
H. Visel
Glossar
Abbreviation PAGCH PAP PBCCH PBX PC PC PCCCH PCCI PCH PCM PCM2FA PCN PCOMP PCU PD PD PDCH PDCP PDH PDN PDP PDTCH PDU PEC PG PH PHF PI PIN PINX PL PLMN PLMN PERMITTED PM PN PNCH PNIC PON POTS PPCH PPE PPF PPP PPSN PR PRA PRACH Meaning Packet Access Grant Channel PPP Authentification Protocol Packet Broadcast Control Channel Private Branch Exchange personal computer Power Control Packet Common Control Channel Physical Communication Channel Identification Paging Channel Pulse Code Modulation Line terminal for adapting S0 bus to 3x64 kbit/s personal communications network Protocol Compression Identifier Packet Control Unit Protocol Discriminator Public Data Packet Data Channel Packet Data Convergence Protocol Plesiochronous Digital Hierarchy Packet Data Network Packet Data Protocol e.g., IP or X.25 Packet Data Traffic Channel Protocol Data Unit Protocol Error Cause Performance Grading PHysical (layer) Packet Handler Function Presentation Indicator Personal Identification Number Private Integrated Services Network Exchange Physical Link Public Land Mobile Network PLMN Permitted for handover purposes phase modulation Pulse Notification Packet Notification Channel (for PTM-M on PCCCH) Pseudo Network Identification Code passive optical network plain old telephone service Packet Paging Channel Primitive Procedure Entity Paging Proceed Flag Point-to-Point Protocol Public Packet Switched Network Path Replacement Primary Rate Access Packet Random Access Channel
Copyright ACTERNA
Page 13
H. Visel
Glossar
Abbreviation PS PS PSD PSDN PSI PSPDN PSS1 PSTN PT PTCCH PTM PTM SC PTM-G PTM-M P-TMSI PTN PTNX PTP P-Type PVC QA QAF QAM QoS R RA RA RAB RAC RACH RADIO-LINK-TIMEOUT RADIUS RAI RAN RANAP RAND RC rDS REC REJ REL REQ RES RES RES RFC RFCH Meaning Pulsed Signal Packet Switched power spectral density Packet Switched Data Network Packet System Information Public Switched Public Data Network Private Signaling System No.1 (=QSIG) Public Switched Telephone Network payload type Packet Timing Advance Control Channel Point To Multipoint Point to Multipoint Service Centre Point to Multipoint - Group Call Point to Multipoint - Multicast Packet TMSI Private Telecommunication Network Private Telecommunication Network Exchange Point To Point Packet Data Type Permanent Virtual Circuit Q-adapter Q - Adapter Function quadrature amplitude modulation Quality of Service Value of Reduction of the MS Transmitted RF Power RAndom mode request information field Routing Area Random Access Burst Routing Area Code Random Access Channel The timeout period for radio link failure Remote Authentication Dial In User Service Routing Area Identity Radio Access Network Radio Access Network Application Protocol RANDom Number (authentification) Rejection Cause remote digital section RECommendation Reject Release Request Response Resume Response Radio Frequency Control Radio Frequency Channel
Copyright ACTERNA
Page 14
H. Visel
Glossar
Abbreviation RFN RLC RLC RLP RLT RNC RNR RNS RNTABLE RNTI RNU ROSE RR RR RR RRC RSC RSE RSVP RT RT RTP RU RXLEV RXLEV-ACCESS-MIN RXLEV-MIN RXLEV-NCELL RXLEV-NCELL-[1-N] RXLEV-SERVING-CELL RXQUAL RXQUAL-SERVING-CELL S S/W S0 S2m SABM SABME SACCH SACCH/C4 SACCH/C8 SACCH/T SACCH/TF SACCH/TH SACK SAGE SAM SANC Meaning Reduced TDMA Frame Number Release Complete Radio Link Control Radio Link Protocol radio line termination Radio Network Controller Receive Not Ready Radio Network Subsystem Table of 128 integers in the hopping sequence Radio Network Temporary Identity radio network unit Remote Operations Service Element Radio Resource Receive Ready Route Restriction Radio Resource Control Reset Circuit Radio System Entity Resource Reservation Protocol Recognition Time Real Time Real Time Transport Protocol Resource unavailable Received Signal Level The minimum received signal level at a MS for access to a cell The minimum received signal level at a MS from an adjacent Received signal level of neighboring (adjacent) cell The received signal level in adjacent cell The received signal level in the serving cell Received Signal Quality Received signal quality of serving cell Supervisory software Basic rate access: 2 B channels (each 64 kbit/s) Primary rate access: 30 B channels (each 64 kbit/s) Set Asynchronous Balanced Mode Set Asynchronous Balanced Mode Extended Slow Associated Control CHannel Slow, SDCCH/4 Associated, Control CHannel Slow, SDCCH/8 Associated, Control CHannel Slow, TCH-Associated, Control CHannel Slow, TCH/F-Associated, Control CHannel Slow, TCH/H-Associated, Control CHannel Selective Acknowledgement Security Algorithms Group of Experts Subsequent Address Message Signaling Area/Network Code
Copyright ACTERNA
Page 15
H. Visel
Glossar
Abbreviation SAP SAPI SB SBC SC SCCP S-CDR SCF SCH SCM SCM SCN SCP SDCCH SDCCH/4 SDCCH/8 SDD SDF SDH SDL SDM SDP SDU SE SEF SEG SEP SF SGM SGSN SH SI SI SIB SIE SIF SIM SIN SIO SIOS SIPO SLA SLS SLTM SM SM SMDS Meaning Service Access Point Service Access Point Identifier Synchronization Burst S-Bus Controller ( So-Chip ) Serial Call Signaling Connection Control Part Serving GPRS Support Node Call Detail Record Service Control Function Synchronization CHannel Signaling Carriage Mechanism subcarrier multiplexing Sub-Channel Number Service Control Point Stand alone Dedicated Control CHannel Stand-alone Dedicated Control CHannel/4 Stand-alone Dedicated Control CHannel/8 space-division duplexing Service Data Function Synchronous Digital Hierarchy Specification Description Language space-division multiplexing Service Data Point Service Data Unit Support Entity Support Entity Function Security Experts Group Signaling End Point Status Field Segmentation Serving GPRS Support Node Satellite Hops Service Indicator Service Interworking Status Indication Busy Status Indication Emergency Alignment Service Information Field Subscriber Identity Module Status Indication Normal Alignment Service Information Octet / Status Indication Out of Alignment Status Indication Out of Service Status Indication Processor Outage Service Level Agreement Signaling Link Selection Signaling Link Test Message Session Management Short Message Switched Multimegabit Data Service
Copyright ACTERNA
Page 16
H. Visel
Glossar
Abbreviation SMS SM-SC SMS-GMSC SMS-IWMSC SN SN SNDC SNDCP SNMP SN-PDU SP SPC SPI SPM SR SRES SRF SRI SRNC SS SS SS #7 SS7 SSF SSF SSP STP S-Type SU SUB SUS SVC TA TA TAC TAF TBC TBF TC TCAP TCH TCH/F TCH/F2.4 TCH/F4.8 TCH/F9.6 TCH/FS TCH/H Meaning Short Message Service Short Message service Service Centre Short Message Service Gateway MSC Short Message Service Interworking MSC Subscriber Number Sequence Number SubNetwork Dependent Convergence SubNetwork Dependent Convergence Protocol Simple Network Management Protocol SNDCP PDU Signaling Point Signaling Point Code Security Parameter Index self-phase modulation Sequence Response Signal RESponse (authentification) Specialized Resources Function Source Routing Information Serving RNC Steady Signal Supplementary Service Signaling System No. 7 Signalling System No. 7 Service Switching Function Sub Service Field Service Switching Point Signaling Transfer Point Signaling Data Type Signal Unit Subaddressing Suspend Switched Virtual Circuit Terminal Adapter Timing Advance Type Approval Code Terminal Adaptation Function Token Bucket Counter Temporary Block Flow Transaction Capabilities Transaction Capabilities Application Part Traffic CHannel A Full rate TCH A Full rate data TCH (<2.4kbit/s) A Full rate data TCH (4.8kbit/s) A Full rate data TCH (9.6kbit/s) A Full rate Speech TCH A Half rate TCH
Copyright ACTERNA
Page 17
H. Visel
Glossar
Abbreviation TCH/H4.8 TCH/HS TCI TCM TCP TDD TDM TDMA T-DT1 TE TEI TEID TFA TFI TFP TFT TIA TID TLLI TMN TMSI TN TOM TOS TP TPS TRAU T-RLC T-RLSD TRX TS TSDI TSPEC TTP TUP TUP+ TXPWR UA UBA UBL UDP UDT UE UEA UI UI UIA Meaning A Half rate data TCH (4.8kbit/s) A Half rate Speech TCH Transceiver Control Interface time-compression multiplexing Transmission Control Protocol Time Division Duplex time-division multiplexing time division multiple access Data Form 1 of the TF Terminal Equipment Terminal Endpoint Identifier Tunnel Endpoint IDentifier Transfer Allowed Temporary Frame Identity Transfer Prohibited Traffic Flow Template Telecommunications Industry Association Tunnel Identifier Temporary Logical Link Identifier Telecommunications Management Network Temporary Mobile Subscriber Identity Timeslot Number Tunnelling Of Messages Type of Service Terminal Portability Three Party Service Transcoder and Rate Adapter Unit Release Complete Released Transceiver Timeslot Transceiver Speech & Data Interface Traffic Specification trail termination point Telephone User Part Telephone User Part plus TX power level in the MS-TXPWR-REQUEST and Unnumbered Acknowledgment Unblocking Acknowledgment Unblocking User Datagram Protocol User Data Transfer User Equipment UMTS Encryption Algorithm Unconfirmed Information Unnumbered Information (Frame) UMTS Integrity Algorithm
Copyright ACTERNA
Page 18
H. Visel
Glossar
Abbreviation UIC Uk2 Uko UL Um UMTS UNI UPA UPD UPI Upo UPT URA USF USIM USR UST UTRA UTRAN UUS V V5TSID VAD VBR VC VCC VCI VCL VCR VDSL VLR VMSC VoD VP VPC VPI VPL VPLMN WAN WAP WDD WDM WS WWW XID XPM Meaning U-Interface Controller ( Uko-Chip ) 4-wire primary rate interface in Germany 2-wire U interface in Germany Uplink GSM/GPRS Air Interface Universal Mobile Telecommunication System User-to-Network Interface User Part Available UP to Date User-Port-Identification 2-wire half-duplex interface (ping-pong ) User Part Test UTRAN Registration Area Update State Flag User Service Identity Module User to User Information User Status UMTS Terrestrial Radio Access UMTS Terrestrial Radio Access Network User-to-User Signaling Variant V5-Timeslot-Identification Voice Activity Detection variable bit rate virtual channel virtual channel connection virtual channel identifier virtual channel link video cassette recorder very-high-speed digital subscriber loop Visitor Location Register Visited MSC, recommendation not to be used] video-on-demand virtual path virtual path connection virtual path identifier virtual path link Visited PLMN wide area network Wireless Application Protocol wavelength-division duplexing wavelength-division multiplexing Work Station World Wide Web eXchange IDentification cross-phase modulation
Copyright ACTERNA
Page 19
H. Visel

GPRS Eng

Uploaded by

Copyright:

Available Formats

GPRS Eng

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

GPRS Eng

Uploaded by

Copyright:

Available Formats

GPRS

General Packet Radio Service

Helmut Visel ACTERNA Arbachtalstr. 6, D-72800 Eningen u.A.

From GSM to GPRS Features and Basic Processes

GPRS Radio Interface

Radio Link Control / Medium Access Control

Protocols Between BSS and SGSN

Layer Model Network Service: Internal Architecture Network Service: Addressing

Communication Between MS and SGSN

The Tunnel Principle and GPRS Tunneling Protocol

10. Data Sheets and Application Notes

Those who know nothing have to believe everything

3 Generations of Mobile Communication 3 Generations of Mobile Communication

Analoge Analoge Mobiltelefonie Mobiltelefonie

Digitale Digitale Mobiltelefonie Mobiltelefonie

Evolution of Bandwidth Evolution of Bandwidth

The Mobile Internet The Mobile Internet

Lower protocol layers

Mobile radio network

Whats In It For Me? Whats In It For Me?

such as the Internet, Intranet or X.25 Networks.

GPRS network X.25 network

e.g. Bluetooth interface

Features and Properties of GPRS Features and Properties of GPRS

GPRS terminals GPRS terminals

Two different types of GPRS mobile stations.

The Old GSM World The Old GSM World

The Old GSM World The Old GSM World

X.25 MSC MSC

Cellular Network Structure Cellular Network Structure

GSM Interfaces GSM Interfaces

Finding the Mobile Subscriber Finding the Mobile Subscriber

VLR TMSI LAI

MSISDN MSRN TMSI LAI

GMSK Modulation GMSK Modulation

MSK modulator MSK modulator

f 124 ... f2 f1 935

915 TS 1 TS 2 TS 3 TS 4 ... TS 1 TS 1 890 t TS 2 TS 2 TS 3 TS 3 TS 4 TS 4 TS 5 TS 5 TS 6 TS 6 TS 7 TS 7 TS 8 TS 8 TS 5 TS 6 TS 7 TS 8

TDMA Frame Hierarchy TDMA Frame Hierarchy

Burst Mask Burst Mask

Tolerance mask 542.8 s (148 bits)

577 s (156.25 bits)

Logical Channels Logical Channels

Speech 22.8 kbit/s

Speech 11.4 kbit/s

Frequency Correction (FCCH)

Random Access (RACH)

Stand Alone (SDCCH)

Data max. 9.6 kbit/s

Data max. 4.8 kbit/s

Broadcast Control (BCCH)

Access Grant (AGCH)

Fast Associated (FACCH)

Slow Associated (SACCH)

Channel Mapping (Downlink) Channel Mapping (Downlink)

Example of a logical channel configuration on the air interface

Channel Mapping (Uplink) Channel Mapping (Uplink)

Example of a logical channel configuration on the air interface

Normal Burst Normal Burst