Tài liệu Điện thoại di động mạng lưới Radio P3 pptx

Table 3.3: The series of the GSM recommendationGSM Public Land Mobile Network PLMN tariffs and international billing procedures secu-rity issues relating to network access, network plann

For this reason, at its general meeting in Vienna in June 1982, CEPT (seeAppendix B.2.2) decided to develop and standardize a Pan-European cellularmobile radio network The aim was for the new system to operate in the

900 MHz frequency band allocated to land mobile radio

A working group, called Group Sp´ecial Mobile (GSM), was set up underthe direction of CEPT There were no guidelines on how the new mobile radiosystem was to transmit analogue or digital speech and data The decision todevelop a digital GSM network was not made until the development stage But

it was agreed from the beginning that the system being planned—called theGSM mobile radio system after the working group that developed it—shouldincorporate and consider new technology from the area of telecommunications,such as ITU-T Signalling System No 7, ISDN and the ISO/OSI referencemodel

Six working groups and three supporting groups were formed to cope withthe enormity of the standardization work The tasks of the different GSMworking groups are listed in Table 3.1

The GSM objectives for its Public Land Mobile Network (PLMN) were tooffer [1]:

• A broad offering of speech and data services

• Compatibility with the wireline networks (ISDN, telephone networks,data networks) using standardized interfaces

• Cross-border system access for all mobile phone users

• Automatic roaming and handover

• Highly efficient use of frequency spectrum

Table 3.1: Tasks of the GSM working groups

proce-dures

mo-bile stations, momo-bile functions and fixed cations networks

Speech Coder Experts

Group (SCEG)

Definition of technique for digitization of speech at

a low bit rateSecurity Experts Group

Support of GSM through satellite systems

Table 3.2: Original timetable for introducing the GSM system

• Support of different types of mobile terminal equipment (e.g., car,portable and hand-held telephones)

• Digital transmission of signalling as well as of user information

• Supplier-independence

• Low costs for infrastructure and terminal equipment

The GSM group tested a number of prototypes for digital cellular radiosystems, and in 1987 decided on a standard that combined the best charac-teristics of different systems A timetable drawn up at the same time for theimplementation of the plan gained the full support of the European Union(EU) (see Table 3.2)

Table 3.3: The series of the GSM recommendation

GSM Public Land Mobile Network (PLMN)

tariffs and international billing procedures

secu-rity issues relating to network access, network planning

station (MS) and base station (BS)

synchro-nization and interleaving

and mobile services switching centre (MSC)

networks and different fixed networks

By 1987, comprehensive guidelines for the new digital mobile radio systemhad already been established by the GSM group By signing the Memoran-dum of Understanding on the Introduction of the Pan-European Digital Mo-bile Communication Service (MoU) on 7 September 1987, the 13 participatingcountries confirmed their commitment to introducing mobile radio based onthe recommendations of the GSM

Later, in March 1989, the GSM working party was taken over by ETSI (seeAppendix B.2.3), and since 1991 has been called the Special Mobile Group(SMG) Today the abbreviation GSM stands for Global System for MobileCommunications, thereby underlining its claim as a worldwide standard

In the meantime all the European countries as well as a large number ofother countries in the world have signed the GSM-MoU agreement and havedeveloped or will be developing mobile radio systems in their countries based

on the GSM recommendations (see Table 3.39)

The planned official start to the GSM system was delayed by one year Onlyfive countries were in a position to undertake test operations on 1st July 1991.The reason for the delay was the level of complexity of the digital network andits components, which is reflected in the voluminous specifications which todaytotal around 8000 pages In 1990 alone another 500 GSM change requests werepassed The entire set of GSM recommendations is divided into 13 series,which cover different aspects of the GSM system, as shown in Table 3.3 (seealso Appendix E)

The GSM recommendations contain detailed specifications for the radio terface which in part are borrowed from the concepts for the analogue nationalcellular standard and ITU-T Rec X.25 However, large parts of the radio in-terface are specific to the GSM system Some of the important features ofGSM include:

in-Frequency band The frequency range between 935 and 960 MHz is used asthe base station transmitting frequency (downlink) and the frequenciesbetween 890 and 915 MHz are used as the base station receiving fre-quency (uplink) The carrier frequencies of the FDM radio channelshave 200 kHz channel spacing in each band, thus providing 124 FDMchannels With time-division multiplexing (TDM), eight communica-tions channels (time slots) are supported per FDM channel

Handover Handover from one base station to another is a mechanism thatallows the connection quality of calls between users to be maintained,interference to be minimized and traffic distribution to be controlled Inaddition, procedures are defined for the re-establishment of a connection

if a handover fails

Power control In the area over 30 dB the equipment of the mobile user and

of the base station controls power in 2 dB steps in order to minimizeinterference

Discontinuous transmission (DTX) GSM offers the option of discontinuoustransmission of speech using voice activity detectors With DTX, trans-mitter battery power is only used when speech or data is being trans-mitted, which minimizes interference and improves the utilization offrequency spectrum

Synchronization Depending on the system, all frequencies and times are chronized with a highly stable (0.005 ppm) reference, which can be cou-pled with a frequency normal

syn-The following features distinquish GSM from other European mobile radiosystems:

• Europe-wide coverage

• Europe-wide standardization

• digital radio transmission

• extensive ISDN compatibility

• protection against eavesdropping

• support of data servicesGSM is regarded as an important advance compared with predecessor sys-tems and is considered to be representative of so-called 2nd-generation sys-tems Along with important technological advances (particularly the intro-duction of digital transmission technology), the standardization of the inter-faces between subsystems in GSM has provided manufacturers and networkoperators flexibility in their development work and configurations

Um A bis

(NSS) Switching Subsystem Network and

(OSS) Subsystem Operation

MS

MS

EIR OMC

AuC HLR

VLR

BSC BTS

BTS

BSC

MSC

BTS MS

Radio Subsystem

O (see below) Points of reference:

Interface to

Transition to

Interface Radio Interface

other Networks

Figure 3.1: Functional architecture of the GSM mobile radio network

In GSM specification 1.02 the GSM system is divided into the following systems [20]:

sub-• Radio subsystem (RSS)

• Network and switching subsystem (NSS) and

• Operation subsystem (OSS)

These subsystems and their components are represented in the simplifiedversion of the functional architecture in Figure 3.1

A GSM mobile station consists of two parts The first part contains allthe hardware and software components relating to the radio interface; thesecond part, known as the subscriber identity module (SIM), stores all thesubscriber’s personal data The SIM is either installed into the terminal orprovided as a smart card, which is about the size of a credit card and has thefunction of a key Once it has been removed from a device, it can only beused for emergency calls, if the network so allows A mobile subscriber canuse the SIM to identify himself over any mobile station in the network, andaccordingly a mobile phone can be personalized using the SIM In addition,each mobile station has its mobile equipment identity (EI).

The following numbers and identities are assigned for the administration

of each mobile station within a GSM network; see Figure 3.57:

• International Mobile Subscriber Identity (IMSI)

• Temporary Mobile Subscriber Identity (TMSI)

• Mobile Station International ISDN Number (MSISDN)

• Mobile Station Roaming Number (MSRN)

Mobile stations can be installed in automobiles or provided as able/hand-portable devices and, according to GSM Rec 2.06, are dividedinto five different classes depending on the allowable transmitter power; seeTable 3.4

port-These classifications also characterize different types of devices: mounted,portable and hand-portable devices Equipment for the GSM-900 class 1 (8–

20 W) has not yet been developed Instead, portable and mounted equipment

is typically found in class 2 (5–8 W) Hand-portable equipment mostly forms with class 4 (0.8–2 W) Class 5 (up to 0.8 W) is also being planned forhand-portable equipment, but places a considerable strain on cellular radiosignal supply This is one of the reasons why it is more suitable for urbanenvironments with small cells, but it is hardly being used anywhere yet An

con-MS can have facilities for both voice as well as data transmission

In addition to the network-dependent radio and protocol functions thatenable access to operation in the network, a mobile station outwardly has atleast one other interface to the mobile subscriber (see Section 3.2.2) It isintended either for a human user (man–machine interface) or for coupling theterminal adapter of another terminal, such as a computer or a fax machine or

Table 3.4: Power classes of mobile stations according to GSM or DCS 1800

• so-called soft keys

Soft keys are function keys used to switch a terminal to different operatingstates They are not assigned a specific function, as is the case with hardkeys, e.g., on a drinks dispenser Consequently the user must be informed ofthe respective function before using the keys

Soft keys are extremely useful with hand-held mobile phones The scriber can use his mobile device with one hand because of the soft key menufunctions that are displayed on the mobile, without having to press key combi-nations at the same time, as is required with the hard key version of a controlpanel

sub-Unlike the conventional telephone, where the user is identified throughthe fixed network connection, radio connections form an anonymous network.Therefore subscriber identification is a prerequisite in a mobile radio networkalone for operational reasons The stored subscriber-related data in a SIMmodule identifies the subscriber when he checks in, and his location area

is derived from the serving base station—an automatic procedure when theterminal is used

In older devices the SIM is installed into the equipment, but the new proach is to plug it in as a card; there are two versions of this:

ap-• smart card, also called standard SIM card

• plug-in SIM card

The only difference between the two cards is their size The standard SIMcard is the size of a credit card based on standard ISO 7816, whereas the plug-

in module is smaller in size and based on the GSM Rec 02.17 [6] In addition

to their size, the cards are also used differently Whereas the standard SIMcard can be activitated simply by being inserted into the card slot provided inthe mobile telephone, the smaller module slides into the equipment mounted

on a cut-down card, which involves first removing the battery The smallerplug-in SIM card has been successful with hand-held mobile telephones.The subscriber-related data is stored in the non-volatile memory of theSIM It can be changed statistically as well as temporarily The permanentdata includes the following elements [6]:

• SIM card type

• IC card identification: serial number of the SIM; identifies card holder

at the same time

• SIM service table: list of additional services subscribed

• IMSI (International mobile subscriber identity)

• PIN (Personal identity number)

• PUK (PIN unblocking key)

• Authentification key Ki

Before a SIM card is assigned to a subscriber, it is first initialized withthis data, and only then can the subscriber use the card to check into thenetwork On the other hand, the dynamic data, which is permanently updatedwhen the terminal is switched on, accelerates the checking-in process becauserelevant information is already stored centrally and there is no need for it to

be requested from the network This includes the following data items [6]:

• Location information: consists of a TMSI, a LAI, a periodically changedlocation updating timer, and update status

• Ciphering key Kc for encoding, and its sequence number

• BCCH information: list of carrier frequencies for cell selection duringhandover and call setup

• List of blocked PLMNs

• HPLMN search: period of time in which an MS roams the home networkbefore it tries to check into another network

Other optional data items can be found in [6] All SIM data is copied in thememory of the MS only for the duration of the active operating state and thendeleted Manufacturers of mobile terminals have the option of additionallyproviding intermediate storage of less important data, such as short messagesand the last-called telephone number However, this data can only be called

up if the equipment is turned on again with the same SIM card that was usedfor its previous deactivation [6]

PIN Except for emergency calls, mobile equipment can only be operated ifthe SIM card has first been activated This is done by the subscriber punching

in a PIN code, which can be between four and eight digits long, after switching

on the equipment When the SIM card is provided by the service provider,the PIN is generally preset with a four-digit number, which the subscriber canchange as often as he likes After the PIN has been correctly entered, thenetwork responds and the mobile is automatically checked in

A PIN can be but should not be disabled, because the subscriber will runthe risk of potential thieves using the mobile free of charge until use of thecard is suspended Anyone who steals an activated mobile phone can only usethe SIM card fraudulently until the first time the equipment is switched off

or the battery runs out If an incorrect PIN is inserted three times in a row,the card will be suspended The subscriber then needs an unlocking key PUK.Some cards are available with a second PIN to protect some of the numbersstored in the card This specifically protects personal telephone numbers andnames entered on the card from unauthorized access The security mecha-nisms and maximum allowable code length of the PIN2 are identical to those

of the PIN [6]

PUK A blocked SIM card can only be released through the use of an PINunblocking key PUK The subscriber is allowed 10 attempts in which to enterthe correct PUK code or else the card will be blocked permanently and can only

be unblocked by the service provider The PUK is an eight-digit permanentnumber that is divulged to the subscriber when he receives the card [6].3.2.1.2 Base Station Subsystem (BSS)

The BSS comprises all the radio-related functions of the GSM network.Depending on the radio transmitting and receiving capabilities of the basetransceiver system, which because of limited transmitter power only suppliescoverage to a specific geographical area within the network, radio cells are cre-ated in which the mobile subscriber is free to roam or communicate The size

of the individual cells depends on a number of parameters, including acteristics of radio wave propagation, local morphology, and expected userdensity in the region

char-A BSS uses transceivers and the following hardware and software to enable

it to connect a mobile subscriber to a number in the public telephone network(PSTN) and allow it to communicate:

• signalling protocols for connection control

• speech codecs (coders/decoders) as well as data-rate adaptation coder/rate adapter unit , TRAU) for access to the network

(trans-• digital signal transmission for coded data

These functions already give an indication of some of the other importanttasks of the BSS Various interfaces have been specified between the BSS andGSM network elements and other networks for the exchange of informationbetween subscribers and the GSM network or other networks; see Figure 3.1.The interface to the mobile subscriber is called the Um-interface It containsspecific parameters for digital radio transmission, such as GMSK modulation,data rate, status of carrier frequencies in the 900 MHz band and channel grid.The BSS is connected to the GSM fixed network over the A-interface (familiarfrom ISDN) with MSCs, the NSS switching centres that provide the subscriberconnectivity to each other and to the external network The A-interface like-wise contains specific digital transmission parameters, including PCM (pulsecode modulation), a 64 kbit/s data rate and a 4 kHz voice bandwidth.Network availability and quality is established by the network operationsand maintenance centre (OMC) of the GSM operator over an O-interface,which provides direct access to BSS units

The elements making up the BSS include:

• Base transceiver station (BTS)

• Base station controller (BSC)

BTS The BTS comprises the transmitting and receiving facilities, includingantennas and all the signalling related to the radio interface Depending on thetype of antenna used, the BTS supplies one or several cells, so, for example,sectorized antennas can supply three cells arranged at 120◦ to each other (seeChapter 2.4)

In a standardized GSM structure the transcoding and rate adaptation unitTRAU is part of the BTS It contains GSM-specific speech coding and decod-ing as well as rate adaptation for data transmission

BSC The BSC is responsible for the management of the radio interfacethrough the BTS, namely for the reservation and the release of radio channels

as well as handover management Its other tasks include paging and mitting connection-related signalling data adapted to the A-interface from/tothe MSC

trans-A BSC generally manages several BTSs, and is linked to the NSS via anMSC

3.2.1.3 Network and Switching Subsystem (NSS)

Switching and network-oriented functions are carried out in a Network andswitching subsystem (NSS) It forms the gateway network between the ra-dio network and the public partner networks (e.g., Public Switched TelephoneNetwork (PSTN), Integrated Services Digital Network (ISDN), Public SwitchedData Network (PSDN)) In their entirety not only are the elements of an NSSpurely physical components but, more importantly, the switching subsystemprovides a large number of functions that are the responsiblity of the manu-facturer and network operator to implement appropriately

The NSS components include the Mobile Services Switching Centre (MSC),the Home Location Register (HLR) and the Visitor Location Register (VLR).Mobile services switching centre (MSC) The MSC is a high-performancedigital switching centre that carries out normal switching tasks and managesthe network Each MSC is usually allocated several base station controllers,and in the geographical area assigned to it carries out the switching betweenmobile radio users and other PLMNs and also forms the link between themobile radio network and the wireline networks (PSTN, ISDN, PDN) TheMSC is responsible for all the signalling required for setting up, terminatingand maintaining connections, carried out in accordance with Common Chan-nel Signalling System No 7, and mobile radio functions such as call reroutingwhen there is strong interference, as part of a handover and the allocationand deallocation of radio channels

Transmission functions for data services are supported through the use

of specific interworking functions (IWF) that are integrated into each MSC.The respective communications channel functions are carried out by facilitiescalled data service units (DSU) The DSU contains functions such as rateadaptation, modem and codec of layer 1, and protocol functions of layer 2.The other tasks of the MSC include the supplementary services familiarfrom ISDN, such as call forwarding, call barring, conference calling and callcharging to the user called The MSC can be envisaged as an ISDN switch-ing centre that has been expanded to include the necessary mobility-relatedswitching functions

Home location register (HLR) All important information (quasi-permanentstatic data) relating to each mobile subscriber, including telephone number,

MS identification number, equipment type, subscription basis and tary services, access priorities and authentication key, is stored in the databasereferred to as the home location register Temporary (dynamic) subscriberdata (e.g., current location area (LA) of the mobile station and mobile stationroaming number (MSRN)) that are necessary for setting up a connection arealso stored When a mobile user leaves his momentary location area (LA),the temporary data held in the HLR is immediately updated The home loca-tion register usually falls under the responsibility of a mobile switching centre

NSS

BSS

Network operation and

Figure 3.2: Structure of an OSS

(MSC) Each mobile subscriber and his related data are registered in only onehome location register in which all the billing and administrative tasks arecarried out In many existent GSM networks there is only one HLR beingimplemented

Visitor location register (VLR) The visitor location register is under thecontrol of an MSC and is used to manage the subscribers who are currentlyroaming in the area under the control of the MSC or, more precisely, in one

of possibly several location areas of the MSC It stores information (e.g., thentication data, international mobile subscriber identity (IMSI), telephonenumber, agreed services) transmitted by the responsible HLR for the mobileusers operating in the area under its control, thereby allowing the MSC tomake a connection The VLR also controls the allocation of roaming numbers(MSRN) to the mobile stations as well as of the TMSI A special dialogue up-dates the VLR if a mobile user moves through several of the MSC’s locationareas The same procedure applies when there is a change of MSC The VLRavoids frequent interrogation of the HLR

au-The functions location area update and call setup and the roles played bythe HLR and the VLR in these functions are described in Sections 3.7 and 3.8.3.2.1.4 Operation Subsystem (OSS)

The operation subsystem in GSM comprises all the important functions for eration and maintenance The user is only indirectly aware of these functionsthrough his experience with a smoothly functioning mobile radio network.The functions of an OSS are allocated to three areas of responsibility (seeFigure 3.2):

op-• Subscription management

• Network operation and maintenance

• Mobile equipment management

The following network elements are part of the OSS:

• Operation and maintenance centre (OMC)

• Authentication centre (AuC)

• Equipment identity register (EIR)

Subscription management Subscription management is able to authenticate

a GSM user from the personal data stored in the HLR (see Section 3.13.1) andprovide him with the agreed services (subscriber data management ) This dataprovides the network operator and the service provider with a call-chargingbasis

Subscriber data management The subscriber data is stored and managed

in the HLR; information relating to data security is in the AuC The HLRcan provide restricted access to elements from other networks, e.g., in order

to allow service providers access to tariff and services data and to ensurethe consistency of data stored in different locations As has already beenmentioned, the SIM card is a dynamically changeable data storage unit duringthe active operation of a mobile station

Call charging Similarly to ISDN, the mobile radio user is charged for servicesused on the basis of so-called call tickets These call tickets are used for billingirrespective of where a call is made in the network The billing location can

be the MSC in which the mobile subscriber is currently active or a gatewayMSC (GMSC) where a communication is connected to an external network.The HLR only stores call-related data Call billing is handled by the re-sponsible OSS subscriber management At the same time tariff data is alsotransmitted between the MSCs or GMSCs and the HLR over the commonchannel signalling system no 7 (SS 7)

Network operation and maintenance The control of network operation andmaintenance tasks uses a separate switching network to connect operatingpersonnel network elements The network is based on the concept of TMN(Telecommunications Management Network ) developed by the ITU-T TheTMN forms an integrated network with its own databases that offer the op-erator options for monitoring, control and intervention

The TMN functions are divided into individual layers similar to the networkelement functions in the ISO/OSI reference model:

Business management Controls the interaction between network and vices and provides information about other service and network devel-opments

ser-Service management Used for the execution of all contractual aspects of aservice between supplier and customer

Network management Supports all network elements and helps to activatefunctions with similar elements of a network.

Network element management Facilitates access to individual network ments

ele-GSM uses standardized concepts for network management, thereby tating the integration of the network elements of different suppliers

facili-The TMN has links with defined interfaces to the network elements of theactive network and to the workstation computers of operating personnel OSSnetwork elements that are connected to several BSS or MSS units are referred

to as OMCs A radio OMC, for example, is responsible for several BSCs andtheir BTSs

Mobile equipment management The management of mobile equipment bythe OSS only concerns information about owner and equipment identity,whereas the MSS coordinates the movements of the equipment, includingroaming, handover and paging For example, an OSS can search for stolen ordefective equipment using its own database, an EIR, for storing data aboutequipment and its ownership (some operators have not established the EIR).Operation and maintenance centre (OMC) The OMC centrally monitorsand controls the other network elements and guarantees the best possible ser-vice quality for a network It relies on services of the network managementand control functions allocated to the network elements by the hierarchicalnetwork management system (TMN) Operator commands are used for inter-vention into the network elements, while the network management is alerted

of any unexpected occurrences in the network The OMC is connected to allnetwork elements over the standardized O-interface (an X.25-interface) Themanagement functions of the OMC include administration of subscribers andequipment, billing, and generation of statistical data on the state and thecapacity utilization of network elements

Authentication centre (AuC) The AuC contains all the information quired to protect a subscriber’s identity, and his mobile communicationagainst eavesdropping, and his right to use the radio interface Because the ra-dio interface is generally susceptible to unauthorized access, special measures(e.g., authentication key assigned to each subscriber and coding of transmit-ted information) were undertaken in order to prevent the fraudulent use ofGSM–PLMN connections Authentication algorithms and encryption codesare stored in the AuC, and strict rules apply for access to this information(see Section 3.13)

re-Equipment identity register (EIR) The EIR is a central database in whichsubscriber and equipment numbers (International Mobile Equipment Identity,

MT2 MT1 MT1

3.2.2.1 User Interface of the Mobile Station

A GSM mobile station consists of the terminal equipment (TE) to which thesubscriber has direct access, a terminal adapter (TA) (if required) and a partthat contains the functions shared by all the services and referred to as mobiletermination (MT) in the GSM specifications The subscriber interface on theterminal (TE) contains the network termination and the different equipmentfunctions (see Figure 3.3)

The following mobile network terminations are used:

MT0 (Mobile Termination Type 0) A network termination for the sion of speech and data integrating the terminal equipment, the terminalequipment functions and sometimes a TA

transmis-MT1 (Mobile Termination Type 1) A network termination with an externalISDN S-interface to which an ISDN terminal (TE1) can be connected

Conventional terminal equipment (TE2) corresponding to the ITU-T,

V or X-series can be connected to an MT1 through the use of an ISDNterminal adapter (TA)

MT2 (Mobile Termination Type 2) This is a network termination with anexternal R-interface to which conventional terminal equipment corre-sponding to the ITU-T, V or X-series can be connected

TE1, TE2 and TA correspond to comparable functional groups of the ISDNconcept The radio interface that supports ISDN-compatible access over trafficand signalling channels is located at reference point Um

3.2.2.2 Radio Interface

This is an important interface in the GSM system, and is therefore covered indetail in Section 3.3

3.2.2.3 BTS–BSC Interface at Reference Point Abis

Transmission over the Abis-interface (see Figure 3.1) is based on PCM-30 and

64 kbit/s interfaces

Because PLMN network operators frequently are not also the operators

of the telecommunications networks, a submultiplex technique that transmitsfour 16 kbit/s channels over a 64 kbit/s channel was standardized to save online costs

3.2.2.4 BSS–MSC Interface at Reference Point A

Speech and data are transmitted digitally over the A-interface (see Figure 3.1),over PCM-30 systems based on the ISDN standard (ITU-T-Series G.732) APCM-30 system has 30 full-duplex channels at 64 kbit/s, with a transmis-sion rate of 2.048 Mbit/s full-duplex Two channels each with 64 kbit/s arerequired for synchronization and signalling (D2-channel)

3.2.2.5 BSC/MSC–OMC Interface at Reference Point O

The O-interface is based on ITU-T recommendation X.25, which was specifiedfor the attachment of data terminal equipment to packet-switched networks.Physically this interface can be implemented over a 64 kbit/s channel Theoption exists to use interfaces of line-switched networks, e.g., V.24bis or X.21

This radio interface is located between the mobile station (MS) and the rest

of the GSM network Physically the information flow takes place between themobile station and the base transceiver station (BTS) But, viewed logically,

7 n+1

4.615

Figure 3.4: Realization of physical channels using FDM and TDM

the mobile stations are communicating with the base station controller (BSC)and the mobile switching centre (MSC) The gross transmission rate over theradio interface is 270.833 kbit/s

Along with voice coding and modulation, multiplexing is also very important

In the GSM recommendations a combination of frequency-division ing (FDM) and time-division multiplexing (TDM) has been standardized, pro-viding multiple access by mobile stations to these systems (FDMA, TDMA).Figure 3.4 shows how a physical channel is produced through a combination

multiplex-of FDM and TDM (see channel 0 on frequency Fn+1 and Sections 3.3.1.1and 3.3.1.2)

GSM utilizes the cellular concept, already proven successful in analoguemobile radio networks, in which a geographical area is divided into plannedradio cells (in the simplest case hexagons), with one BTS per cell with whichthe mobile stations can make contact The radio cells, each having the exclu-sive use of specific FDM channels, are combined into groups (clusters) Thesame frequencies are only reused after a sufficiently long distance in neigh-bouring clusters (see Section 2.3)

The cell radius can vary according to user density The likelihood that amobile user will leave a cell during a call, thereby necessitating a handover,

is less in large radio cells than in small cells Small cells, on the other hand,make more efficient use of a frequency band because they operate with a lowertransmitter power, the cluster is less spread out and consequently the available

200 kHz

124 3

2 1 Channels:

935 MHz

890

Frequency Band of the Mobile Station Frequency Band of the Base Station

Figure 3.5: Frequency bands used by GSM

frequencies can be reused at smaller physical intervals In practice, the size ofcells is determined by traffic volume, the maximum transmitter power of theBTS of the frequencies allocated to a cell and morphological conditions.Thus cells in rural areas can have a radius of up to 35 km Larger cellradii would cause a higher round-trip propagation delay; the maximum delay

is 0.233 ms, much larger than specified in the standard In metropolitan areasthe radius might only be at 300 m, which allows a traffic volume of up to

200 Erl./km2 Cells are divided into sectors in order to increase capacity (seeSection 2.4)

3.3.1.1 Frequency-Multiplexing Structure

One of the most important criteria in designing a radio interface was efficientutilization of the available frequency band In Europe two 25 MHz wide fre-quency bands in the 900 MHz band were reserved for GSM Transmissionfrom the mobile unit to the base station (uplink) takes place in the 890 MHz

to 915 MHz range; in the reverse direction (downlink) the 935–960 MHz quency band is used in a frequency-division duplex (FDD) mode of operation

fre-15 MHz at the lower band limit and 1 MHz at the upper band limit will not beavailable until 2001 After current use is discontinued, an additional 10 MHzbetween 880 and 890 MHz and between 925 and 935 MHz will be available as

a GSM extension band (see Appendix C) A duplex interval of 45 MHz existsbetween the transmit and receive frequencies

The frequency bands are divided into 200 kHz bandwidth channels, fore providing a total of 124 FDM channels each for transmitting and receivingoperations (see Figure 3.5)

there-Each mobile station can occupy all 124 carrier frequency pairs, althoughaccording to the GSM specifications use of channels 1 and 124 should beavoided if possible The respective 200 kHz bandwidth is kept as a guard bandfor the neighbouring systems in the frequency band If the carrier frequencies

on the uplink are denoted by Fu and those on the downlink as Fd then theGSM band can be defined as

F (n) = 890.2 MHz + 0.2(n− 1) MHz (1 ≤ n ≤ 124) (3.1)

0 1 2 3 4 5 6 7

time slot 156.25 bit 0.577 ms burst 148 bit

4.615 ms

time slot:

Figure 3.6: Structure of a TDMA frame

Fd(n) = 935.2 MHz + 0.2(n− 1) MHz (1 ≤ n ≤ 124) (3.2)and the extension band as

in multiple access, the frame is referred to as TDMA frame in the GSM ommendations

rec-A physical channel is characterized by its carrier frequency and the timeslot available to it, which recurs every 4.615 ms Each time slot has a lengthcorresponding to the duration of 156.25 bits or 0.577 ms (15/26 ms) Thislength is produced from the transmission rate of the modulation method(1625/6 kbit/s) and the number of bits to be transmitted in a slot A slot isused by a burst with a length of 148 bits, which, corresponding to the guardtime, is 8.25 bits shorter in duration than the slots to avoid overlapping withother bursts Data is transmitted in bursts If messages are longer than aburst, they are split up among several bursts and then transmitted

Overall there are five types of bursts (see Figure 3.7 [14]) which differ fromone another in function and content The tail bits that occur in all burstsare defined as modulation bits and always have the same value as specified inthe standard The bursts are sent so that the bits with the lowest value aretransmitted first

Normal burst For transmitting messages in traffic and control channels

TB 3 Guard

8.25 Guard TB 3

Bits

TB 3 Guard Encrypted

Training Sequence Extended

Encrypted Bits Sync.-Sequence

Training 26

57 57

58

41

8.25

Figure 3.7: Bursts used in GSM

Figure 3.8: Envelope of the radio signal of a burst

Access burst Used for call setup This burst is shorter than the others cause it does not require the MS to be fully synchronous with the BTS.Synchronization burst Sent by the base station and used for synchronization.Frequency correction burst Sent by the base station and used for frequencycorrection at the mobile station to prevent possible interference fromneighbouring frequencies

be-Dummy burst Placed in an empty slot if no data is being sent

The signalling characteristics of a burst over time are not allowed to exceedthe area of a prescribed mask (see Figure 3.8) In the area of the tail bits andthe guard space the signal can deviate considerably from the standard 0 dB It

is clear that neighbouring bursts only minimally overlap in the same TDMAframe

iˆjLTWlxehv&ynWo ikNOnWo‰VŠMqRr

Figure 3.9: Time delay between uplink and downlink

The time-division multiplexing technique is applied to the uplink and tothe downlink channel So that the mobile stations do not have to transmit andreceive at the same time, the TDMA frames from the uplink are transmittedwith a delay of 3 time slots (see Figure 3.9) The parameter timing advance(TA) is used by the BTS to compensate for the round-trip signal propagationdelay BTS-MS-BTS The value of the 6 bit of TA indicates to a receivingmobile how many bit durations (3.7 µs each) it must transmit its burst earlierthan as derived from the received slot tact signal to reach synchronizationwith the slot tact defined by the BTS

Since multipath reception and co-channel interference can affect the quality

of certain FDM channels, an optional method called frequency hopping isapplied With this method the frequency is changed after each transmittedframe of a channel (see Figure 3.10) The frequency change, which can lastapproximately 1 ms, takes place between the receiving or the transmittingtime slots

The sequence of frequencies in a hopping cycle through which a mobile tion passes is calculated with an algorithm implemented in each MS Theadvantage of this procedure is that all mobile subscribers are guaranteedtransmission channels with nearly the same quality During data transmis-sion, interference from co-channels in the cycle is limited for each frequency

K1 K1 K1

K1

K1

Physical Channel with Data Rate 4a

with Data Rate 3

Logical Channel K1

with Data Rate Logical Channel K2

Figure 3.11: Relationship between logical and physical channels

to the duration of one burst only and can be eliminated through error dling; effective error-correction procedures are standardized for voice and datatransmission

Logical channels occur through the allocation of time slots by physical nels Consequently the data of a logical channel is transmitted in the cor-responding time slots of the physical channel During this process, logicalchannels can occupy a part of the physical channel or even the entire channel.For instance, if a physical channel has a transmission rate of 4a, then a logicalchannel K1 with a data rate of 3a and a second logical channel K2 with adata rate a can transmit on the same physical channel (see Figure 3.11).The GSM recommendations define several logical channels for signalling

chan-on the basis of this principle, dividing them into two main groups: trafficchannels and control channels

Table 3.5: Traffic channels in the GSM recommendation

3.3.3.1 Traffic Channels

Traffic channels (TCH) are logical channels over which user information areexchanged between mobile users during a connection Speech and data aredigitally transmitted on these channels using different coding methods.Different transmission capacities are required depending on the type ofservice used (e.g., voice transmission, short-message service, data transfer,facsimile) A distinction is therefore made between the following traffic chan-nels:

Bm-channel Transmission over a Bm-channel (m=mobile), which is also called

a full-rate traffic channel (full-rate TCH ), is carried out at a gross datarate of 22.8 kbit/s Digitalized and coded speech only require 13 kbit/sfor transmitting voice information The remaining capacity in voicetransmission is used for error correction It is possible to transmit data

at 12, 6 or 3.6 kbit/s over a Bm-channel

Lm-channel The half-rate traffic channel (half-rate TCH ) transmits at a grossrate of 11.4 kbit/s The number of channels in GSM can be doubled

in a given frequency band because of the speech codecs available forhalf-rate channels Efficient speech coding algorithms were developed in1995; they were introduced commercially in 1997/98 Half-rate TCHsallow data to be transmitted at bit rates of 6 or 3.6 kbit/s

Table 3.5 lists the traffic channels specified in the GSM recommendation.3.3.3.2 Control Channels

Control information is used for signalling and for system control and is notpassed down to the subscribers Typical signalling tasks include the signallingfor establishing, maintaining and releasing traffic channels, for mobility man-agement and access control to radio channels

Control information is transmitted over so-called control channels (CCH),which, following ISDN, are also referred to as Dm-channels The control chan-nels offer the mobile stations a packet-oriented continuous signalling service

Table 3.6: Control channels in GSM

enabling them within the PLMN to receive messages from the base stationsand to send messages to the base stations at any time

Because the control and management of a mobile radio network is far morecomplex from the standpoint of signalling than a fixed network, three groups

of control channels were defined in GSM:

• Broadcast control channel (BCCH)

• Common control channel (CCCH)

• Dedicated control channel (DCCH)

Table 3.6 contains a list of all the control channels defined in the GSM ommendations, and in the directional column indicates the directions possible

rec-on each channel (uplink, downlink or both)

Broadcast Control Channel (BCCH) This channel is used to transmit formation about the PLMN from the base station to the mobile stations inthe radio cell through a point-to-multipoint connection The kind of informa-tion conveyed over a BCCH includes identification of the network, availability

in-of certain options such as frequency hopping and voice activity detection andidentification of the frequencies being used by the base station and neighbour-ing base stations

One of the subchannels of the BCCH is the frequency correction channel(FCCH), used for transmitting a frequency correction burst to the mobilestation for possibe correction of the transmitting frequency

Another subchannel of the BCCH is the synchronization channel (SCH),used for transmitting synchronization bursts to a mobile station to allow it totime-synchronize

Messages transmitted over the BCCH and its subchannels are transmittedexclusively in simplex mode by the base station to the terminal equipment

Common control channel (CCCH) This designation is an umbrella termfor control channels that handle the communication between the network andthe mobile phone Included among the CCCH channels are:

Paging channel (PCH) This channel exists only on the downlink, and is tivated for the selective addressing of a called mobile terminal during aconnect request from the network (incoming call)

ac-Random access channel (RACH) This access channel only occurs on the link, and allows the mobile station, using an S-ALOHA access protocol,

up-to request channel capacity from the base station up-to establish a tion

connec-Access grant channel (AGCH) The base station uses this logical channel torespond to a message received over the RACH from a mobile station

In accordance with the call setup mechanism selected by the networkoperator, the mobile station is allocated an SDCCH or a TCH over theAGCH that only exists on the downlink; see Section 3.5.1

Dedicated control channel (DCCH) This designation is an umbrella termfor three bidirectional point-to-point control channels that are used to trans-mit signalling messages for call control at different bit rates The three DCCHchannels are:

Stand-alone dedicated control channel (SDCCH) This channel is alwaysused when a traffic channel has not been assigned, and is allocated

to a mobile station only as long as control information is being mitted The channel capacity available from an SDCCH is 782 bit/s,which is much lower than that of a TCH Control information transmit-ted on the SDCCH includes registration, authentication, location areaupdating and data for call setup

trans-Slow associated dedicated control channel (SACCH) This channel is ways allocated parallel to a TCH or an SDCCH It is used to transmit

al-at a dal-ata ral-ate of 383 bit/s system informal-ation from the network to themobile station and measurement data on signal strength and receivequality from the MS to the network

Fast associated dedicated control channel (FACCH) This channel is set up

in the short term only when a traffic channel exists and then it usesits time slots As an example, an FACCH is set up for an impendinghandover and the necessary control data is transmitted over the FACCH.This channel can handle bit rates of 4600 bit/s or 9200 bit/s

3.3.4 Hierarchy of Frame Structures

The division of the GSM frequency spectrum into 124 FDM channels and itsTDMA channel structure, the tasks of the logical channels and the use of thedifferent bursts have been explained in previous sections

In GSM the TDMA frames, which contain eight time slots for ting bursts, are combined into multiframes A distinction is made betweenmultiframes of two different lengths:

transmit-• 26-frame multiframe • 51-frame multiframe

The bursts of the traffic channels (TCH) and the SACCHs and FACCHsassigned to them are transmitted in 26-frame multiframes Each traffic chan-nel is allocated 1 of 8 (with half-rate transmission, 16) time slots of a TDMAframe The associated 8 SACCHs are transmitted in the 12th TDMA frame.The last TDMA frame (25) of the 26-frame multiframe is only used if another

8 SACCHs are required for half-rate transmission Time slots are stolen fromthe TCHs for the transmission of the FACCHs

The data of the FCCH, SCH, BCCH, RACH, AGCH, PCH, SDCCH,SACCH/C and CBCH channels is sent in 51-frame multiframes It was gen-erally specified that speech and data would be transmitted in 26-frame mul-tiframes and signalling data (except for SACCH/T and FACCH) in 51-framemultiframes However, there was some deviation from this rule when thepacket-data service was introduced (see Section 3.11.3)

Fifty-one of the 26-frame multiframes and 26 of the 51-frame multiframesare combined into a superframe, and 2048 superframes produce a hyperframe(see Figure 3.12) It takes almost 3.5 hours to transmit a hyperframe.Figure 3.13 shows the relationship between 51-frame and 26-frame multi-frames The BCCH frames in a 51-frame multiframe are decoded every 240 msduring the unused 26th frame of the 26-frame multiframe

Logical channels are mapped onto a physical channel through a process ofcyclically recurring multiframe patterns The correct positioning timewise

of the cycles is achieved through synchronization of BTS and MS The BSSprovides a reference counter in each cell that is used to number the time slotsand serves as the mapping scheme for the multiframes In GSM each timeslot is given a number This number along with the TDMA frame numberallows each time slot to be clearly identified (see Figure 3.9) Because of aframe structure encompassing several levels, the counter for the TDMA framenumbers can go as high as 2 715 648 This high a number was necessary forcryptographic reasons, because the enciphering algorithm uses the TDMAframe number as an input parameter

TCH, SACCH/T, FACCH

1 Superframe = 1326 TDMA Frames (6.12 s)

1 Hyperframe = 2048 Superframes = 2 715 648 TDMA Frames (3 h, 28 min, 53 s, 750 ms)

Figure 3.12: The structure of a transmission medium with TDMA frames, frames, superframes and hyperframes

multi-Figure 3.13: The relationship between 51-frame and 26-frame multiframes

Figure 3.14 shows how traffic and signalling channels are mapped onto aphysical (frequency) channel For each TDMA frame of the correspondingmultiple frame, a slice of the ‘cylinder’ shown is provided that carries eighttime slots (the TDMA frame) [33]

B B

B B

B B B

12 13 14 15

6 7 8 9

2 3 4 5

46 47 48 49

B

50 25

24 23 22 21

0 1

3 2 4 5 6 7 8 9 10 11 12

14 13 15 17 18 19 20

19 18 17 16

0 1

11

0 10

20 16

10

50

A A

2

2 5

Figure 3.14: Traffic and control channel with ‘wind up’ time axis

T T T T T T T T T T T T A T T T T T T T T T T T T

26-frame Multiframes

Downlink and Uplink (Physical Channel with Even Time Slot Number)

Downlink and Uplink (Physical Channel with Odd Time Slot Number)

T: TCH/F

A: SACCH/TF

Figure 3.15: Channel combination 1: TCH/F + FACCH/F + SACCH/TF

3.3.5.1 Approved Channel Combinations

Logical channels are not simply distributed over the two multiframes, butfollow a particular pattern The following is a set of approved channel com-binations (CC) [14], represented by their multiframes:

The shared use of a physical channel by several logical channel types doesnot mean that all of these channels can be used at the same time Althoughthe mapping of combinations of logical channels onto a physical channel maymean that several logical channels can appear on one physical channel, theyoccur sequentially at intervals of at least one TDMA frame length The se-quential numbering of the TDMA frames ensures that the physical layers ofboth communicating partners use the currently appropriate logical channelfor transmitting and receiving

The multiframes in the approved channel combinations are shown in ures 3.15–3.21 The mapped multiframe combinations are derived from tablesfound in Series 05.02 of the GSM standard [14] The tables describe whichtime slots and which frequencies the respective logical channel is allowed touse

Fig-T0 T1 Fig-T0 T1 T0 T1 T0 T1 T0 T1 T0 T1 A0 T0 T1 T0 T1 T0 T1 T0 T1 T0 T1 T0 T1 A1 Downlink and Uplink

T0: TCH/H(0)

T1: TCH/H(1)

A0: SACCH/TH(0)

Figure 3.16: Channel combination 2: TCH/H(0,1) + FACCH/H(0,1) +

SACCH/TH(0,1)

T0 T1 T0 T1 T0 T1 T0 T1 T0 T1 T0 T1 T0 T1 T0 T1 T0 T1 T0 T1 T0 T1 T0 T1 A0 T0 T1 T0 T1 T0 T1 T0 T1 T0 T1 T0 T1 A0 T0 T1 T0 T1 T0 T1 T0 T1 T0 T1 T0 T1

T0 : TCH/H(0)

Downlink and Uplink (Physical Channel with Even Time Slot Number)

Downlink and Uplink (Physical Channel with Odd Time Slot Number)

Figure 3.17: Channel combination 3: TCH/H(0) + FACCH/H(0) +

R: RACH

51-frame MultiframesFigure 3.19: Channel combination 5: FCCH + SCH + BCCH + CCCH +SDCCH/4 (0, 1, 2, 3)

A0: SACCH/C8(0) A1: SACCH/C8(1) A2: SACCH/C8(2) A3: SACCH/C8(3)

A4: SACCH/C8(4) A5: SACCH/C8(5) A6: SACCH/C8(6) A7: SACCH/C8(7) 51-frame MultiframesFigure 3.21: Channel combination 7: SDCCH/8 + SACCH/8

A cell in a GSM-PLMN has a set of n + 1 frequency channels available forits use These frequency channels are denoted by C0, C1, , Cn The indiceshave no relationship to the frequency The described channel combinationsare not allowed to occupy just any random time slots or frequencies

For example, the combination 4 shown in Figure 3.18 is only allowed to betransmitted in time slot 0 of the carrier frequency C0 On the downlink theBTS must transmit in each time slot number 0 the logical channel of combi-nation 4 to enable the mobile station to carry out performance measurements

If no information is available for sending, a dummy burst is transmitted (seeSection 3.3.1.1)

The combination 5 shown in Figure 3.19 is only allowed to occupy timeslot 0 in frequency C0 None of the other combinations, except for combina-tions 4 and 5, are allowed to use this physical channel

The channel combination 6 (see Figure 3.20) occupies time slots 2, 4 and

6 of the carrier frequency C0 Except for the restriction mentioned above,channel combinations 1, 2, 3 and 7 are allowed to use any time slots in anyfrequencies

3.3.6 Channel Combinations of a Cell Depending on

Anticipated Cell Utilization

The BSS provides a cell with a set of logical channels occupying several cal channels On the basis of the anticipated traffic load of a cell, the networkoperator establishes a channel configuration that must adhere to the rulesmentioned in the last section Each individual transceiver (TRX) can of-fer eight channel combinations in each time slot The time slot is identifiedthrough a time slot number (TN) Three common combinations are brieflypresented here:

physi-• Low-capacity cell with one TRX:

– TN 0: FCCH + SCH + BCCH + CCCH + SDCCH/4(0,1,2,3)+ SACCH/C4(0,1,2,3)

– TN 1 to 7: TCH/F + FACCH/F + SACCH/TF

• Medium-capacity cell with four TRXs:

– Once on TN 0: FCCH + SCH + BCCH + CCCH

– Twice (on TN 2 and TN 4): SDCCH/8 + SACCH/8

– 29 times: TCH/F + FACCH/F + SACCH/TF

• High-capacity cell with 12 TRXs:

– 87 times: TCH/F + FACCH/F + SACCH/TF

After successful synchronization, the mobile station is informed throughthe system information of the BCCH which channel combinations on whichphysical channels are being offered to it by the BSS Depending on its currentoperating state (idle state or dedicated state), it uses a particular subset fromthis offering of channels

What is noticeable is that a BCCH always appears together with the logicalchannels SCH und FCCH and that it can always be found in time slot 0.This condition helps to facilitate synchronization when an MS makes its firstcontact with a BTS Additional combinations 6 (see Figure 3.20) are addedwhen traffic is expected to be heavy

Primitives

MPH-Radio Resource Management (RR)

RR-Primitives

PH-Primitives Physical Layer

Data Link Layer

Figure 3.22: Interfaces of layer 1 (physical layer )

GSM layer 1 is essentially comparable to the physical transmission layer ical layer ) in the ISO/OSI reference model (see Section 2.5) Unlike theISO/OSI model, an entity of layer 3, radio resource management (RR), di-rectly accesses this lowest layer (see Figure 3.22)

(phys-This access allows channel allocations to be controlled and layer-3 tion to be queried on the state of the physical layer and the radio connection,such as channel values, error ratios and received signal strength This in-formation is required by layer 3 in order for it to fulfil the typical cellularnetworks tasks such as handover and roaming

informa-Layer 1 of a GSM implementation is responsible for the radio transmission

of traffic and signalling information Its main tasks include:

• Producing bursts, multiplexing the bursts in TDMA frames, and mitting the frames on the available physical channels over dedicatedcontrol and traffic channels

trans-• Seeking and seizing a broadcast control channel (BCCH) by the MS

• Judging channel quality and received signal strength

• Error detection and correction mechanisms (defective blocks are notforwarded to layer 2)

• Synchronization with frame transmission

• Encoding of data stream

The GSM system uses a digital 0.3 GMSK modulation format GMSKstands for Gaussian Minimum-Shift Keying, in which minimum shift meansthat there is no discontinuity between the phase of the carrier and the time.The modulation format 0.3 GMSK means that the modulated signal passesthrough a Gauss filter with the product of the 3 dB bandwidth B and the bitperiod T producing the value BT = 0.3

Table 3.8: Coding methods for logical channels

3.3.7.1 Channel Coding and Interleaving

This section describes how GSM layer-2 messages are conveyed by means

of bursts The two dedicated signalling channels SDCCH and FACCH areespecially important In [12] they are referred to as the main signalling link.Table 3.8 presents a general overview of the coding procedures used for logicalchannels The following information is provided for each type of channel:

• The number of bits per 456 bit-block of user information which aresystematic redundancy and which are filler bits

• The coding rate of the convolutional code used

• The depth of bit interleaving

Different channel coding methods are applied on the various channels forthe different services (see Section 2.7.1)

A Fire code—a linear, binary block code family derived from the cyclicblock code—is used with most signalling channels This code has a generatorpolynomial which offers good error detection and/or error correction whenerrors occur in groups (burst errors):

(X23+ 1)(X17+ X3+ 1)Forty Fire code parity bits are added to the 184 data bits of a layer-2protocol data unit (see Figures 3.23 and 3.24) This allows groups of errors

of up to 11 bits to be detected and corrected When layer 1 recognizes adefective block and can no longer correct the errors, it does not forward thisblock to layer 2

D: SDCCH Segmentation in eight Subblocks

FEC through Fire Code, Tail Bits

Figure 3.23: Coding and interleaving in an SDCCH

All channels have forward error correction (FEC) protection through volutional coding The respective code ratios are given in Table 3.8 A con-volutional code with the ratio r = 1/2 and the influence length k = 5 (seeSection 2.7.3.2) is used on the control channels

con-Interleaving allows correlated errors caused by a short signal fading to pear as single errors at the decoder of the receiver (see Figure 3.25)

ap-Two types of interleaving are in use, bit interleaving and block interleaving.The bit interleaving with a depth of n is based on an algorithm, where the 456information bits of a convolutional coded layer-2 block (see Figure 3.23) arearranged into rows mod n into a table (see left-hand upper side of Figure 3.25).The bits are read out for transmission from the table row by row (seeright-hand upper side and lower side of Figure 3.25)

During this bit interleaving procedure, a block interleaving is introduced inaddition whereby the first four rows of 57 times 8 bit (right-hand upper side

of Figure 3.25), which each fill half a normal burst of a TCH, are inserted intobit positions in this burst with even bit numbers The remaining four rowsare transmitted on odd bit positions in the normal burst Each of these rowscorresponds to one subblock shown in Figure 3.23

The de-interleaver at the receiver rearranges the interleaved bits into theoriginal sequence It is clear from the algorithm that a group of errors ofmaximum length n− 1 applied to the interleaved bit stream on the radiomedium is resolved into single bit errors after de-interleaving

The block interleaving procedure further contributes to spread originallyneighbouring bits during the radio transmission to improve the resistanceagainst the so-called error bursts resulting from signal fading

Figure 3.26 (b) shows the error protection mechanisms for the Random cess channel (RACH); see also Figure 3.32 Because a mobile station’s callrequest should be recognized quickly by the base station and collisions cansometimes occur that must quickly be resolved, the 8-bit message of the ac-cessing mobile station (code for type of access and random number along with

Layer-2 Block Segmentation in Eight Subblocks

455 454 453 452 451 450 449 448 447 446

0 0 1 1 0


 
   
 
  ! "

5 4 3

2

1

13 12 11

3 11 19 18 17 16

4 12 20

5 13

6 14 21 22

441 442 443 444 445 446 447

4 3

Rearrangement

Parity bits

Tail bits

Random number Cause and

456 bit / 20 ms = 22.8 kbit/s

260 bit / 20 ms = 13.0 kbit/s

Figure 3.26: Channel coding for (a) speech transmission and (b) signalling

check sum) is modified through the base station identification code (BSIC),then convolutionally coded to 36 bits and transmitted without interleaving

On the FACCHs block diagonal interleaving occurs in eight half bursts; onall other control channels (with the exception of RACH) interleaving is in fourwhole bursts This is illustrated on the basis of an SDCCH (see Figure 3.23)and an FACCH/F (see Figure 3.24)

For example, at a transmission rate of 9.2 kbit/s, every 20 ms 184 bitsare transmitted on an FACCH For the detection of non-correctable errors

40 parity check bits calculated according to the Fire code are added by thephysical layer All bits are then protected by a convolutional code of ratio

r = 1/2 This provides sufficient error correction protection on the receiverside The resulting 456 bits (= 4· 117) are then distributed over the four timeslots of a logical channel An interleaving of 4 is continuously applied to thesignalling channels because messages are too short and otherwise too manyslots would be occupied.

3.3.7.2 Speech Coding for Radio Transmission

In line with GSM goals on frequency economy, the transmission rate was not

to exceed a value of approx 16 kbit/s The speech codec RPE-LTP wasselected It is based on the LPC technique (linear predictive coding) andincorporates a combination of long-term prediction (LTP) and regular pulseexcitation (RPE) The codec has a net bit rate of 13.0 kbit/s Togetherwith the redundancy for error protection, 22.8 kbit/s (gross bit rate) aretransmitted on the radio channel Speech is produced in packets of 20 ms eachfrom the speech codec With consideration given to the channel characteristicsand the possible bit error ratio of the speech codec parameters, the followingthree-stage channel coding method was selected (see Figure 3.26 (a)):Ordering of the bits according to importance The bits of the speech codecparameters are arranged in descending order of importance, with themost important bit, No 1, first and the least important bit, No 260,last

Cyclic redundancy check (CRC) for the most important 50 bits (Class1a) These are extended by three parity bits as a cyclic redundancycheck (CRC) The CRC bits are inserted after bit No 50, and the bitsthat follow are renumbered The associated generator polynomial isg(x) = x3+ x + 1 and is implemented through a cyclic shift registerwith appropriate taps On the receiving side these CRC bits are usedfor error detection and can sometimes indicate one or more errors withinthe most important bits

Convolutional coding of the most important 185 bits (Class 1a+1b) forerror correction Four tail bits (0) are attached to the first 185 bits, andthese 189 bits are then recoded to 378 bits using a convolutional cod-ing with ratio r =1/2 Since this protects the first and last bits betterthan it does the bits in the middle of a sequence, the first 185 bits arerearranged further so that the most important bit ends up first, thesecond most important bit is last (185), the third most important bit

is second, and so forth The remaining 78 bits of the speech codec rameters (Class 2) are transmitted without protection so that altogether

pa-456 bits are produced per speech coding frame and transmitted with theinterleaving factor 8

With interleaving (distribution of data over several time slots of a logicalchannel), transmission errors occurring in bursts are distributed more

evenly over the speech data received, and transmission errors occur assingle bit errors and therefore can be corrected.

Normal bursts (see Figure 3.7) with speech data are divided into two arate blocks, each with 57 bits and one control bit, which transport data ofdifferent (consecutive) speech coder frames so that eight sequential time slotscontain the data from a speech coder frame of 20 ms duration and in additionthe same amount of data from a second speech coder frame Four consecutivetime slots carry 456 user bits in 18.5 ms, namely the data output generatedevery 20 ms by the speech codec The control bit indicates whether a block iscarrying user data or data from the fast-associated control channel (FACCH).3.3.7.3 Enhanced full rate (EFR) speech coder

sep-The EFR speech coder is an Algebraic Codebook Excited Linear Predictive(ACELP) coder that has been introduced into products since 1998

It produces 244 code bits for every 20 ms long speech frame, thereby responding to a bit rate of 12.2 kbit/s These code bits are divided into threemain categories upon their relative subjective importance The most impor-tant 65 bits are protected by a half-rate convolutional encoder and an eight-bitcyclic redundancy checksum If any of these bits are in error, the decoder willreject the speech frame as being in error The following 117 bits are protected

cor-by the same half-rate convolutional code, whereas the remaining 78 bits arenot offered any protection at all

3.3.7.4 Discontinuous Voice Transmission

As an option, GSM permits the use of discontinuous transmission (DTX)for the transmission of voice signals, thereby reducing the amount of powerhandsets need in order to transmit and co-channel interference Voice pausesare recognized by a function referred to as voice activity detection (VAD)and transmission is halted until data is again available The traffic channelremains assigned to the connection

To give the receiver the impression that a connection still exists ing transmission pauses, the receiver side produces something called comfortnoise, which is provided by the measurement results of the previous voicetransmission phase and is activated by the transmission of updated measure-ment values from the transmitter during long voice pauses

dur-3.3.7.5 Adaptive Equalization of the Radio Signal

The 26 bits in the middle of each time slot (training sequence) are used bythe receiver in order to adjust the parameters of its matched filter and to beable to select the path with the strongest signal in multipath propagation ofthe radio waves in its equalizer

In the GSM frequency band, shadowing may happen to the radio signals

of mobile and base stations; this is partly compensated by the fact that radio

In the GSM frequency band, shadowing may happen to the radio signals

of mobile and base stations; this is partly compensated by the fact that radio. .. RPE-LTP wasselected It is based on the LPC technique (linear predictive coding) andincorporates a combination of long-term prediction (LTP) and regular pulseexcitation (RPE) The codec has a net... calculated according to the Fire code are added by thephysical layer All bits are then protected by a convolutional code of ratio