Fundamentals: Signalling at the Air-Interface

Page 1

Fundamentals: Signalling in GSM Signalling is the language used for communication between machines or computers. The language understood by the machines and used for information transfer is described by protocols. The chapter below gives a short overview of the signalling procedures used in GSM. However, it only describes signalling at the air interface and not the whole of the signalling in the backbone network. Tackling the theory first, a description of the OSI reference model will make signalling easier to understand.

The OSI Reference Model A system as complex as GSM requires a lot of planning and organization both at the definition and implementation phase. A structure for a generic data communication network has been developed by the International Standards Organization ISO and is referred to as the open system interconnection (OSI) model. The OSI reference model provides for a number of layers, each layer communicating exclusively, and according to well-defined rules, with the layers immediately above and below it. Communication is, therefore bi-directional except for the lowest, or physical layer, where information passes only to the layer above. Tasks can, therefore, be assigned to specific layers and transactions modularized. General "rules" for the OSI model: The layers operate independently of each other and this also applies to any two consecutive layers in the hierarchy. Each layer can be thought of as providing a service to the layer immediately above and receiving a service from the layer immediately below. Each layer directly communicates only with the layer immediately above by exchanging primitives. The latter are instructions to the protocol layer in question and help to transfer information. Each layer also communicates with its corresponding layer at the remote end. This is called peer-to-peer communication. Layers 4 to 7 are required only in the terminal equipment. In our example, where a call is made by the mobile, layers 4 to 7 are incorporated by the equipment involved (session and transport layer), the presentation of language (presentation layer) and the call content (application layer). The functionality of layers 1 to 3 is available between two network nodes, e.g. between the mobile station and the network. The GSM specifications follow the stipulations for the bottom three layers of the OSI model.

Rohde & Schwarz Training Center V 1.0

Fundamentals: Signalling at the Air-Interface

OSI-Modell

7

Application Layer

6

Presentation Layer

5

Session Layer

4

Transport Layer

3 2 1

Network Layer

Page 2

Realisation in GSM

Tasks of the user

Taks of the fixed network

Call Control Mobility Management Radio Ressource Management

Data Link Layer

Segmentation / Concatenation Acknowledgement

Physical Layer

Tasks of the GSMnetwork

Forward Error Correction Channel Coding Modulation

Figure: OSI reference model as applied to GSM In the lowest layer (Layer 1), the physical characteristics of the transmission medium are specified. In the context of GSM radio links, this definition not only includes the RF carrier frequency and GMSK but also correct timing of burst transmissions necessary because of the use of TDMA, time-division multiple access. This layer also incorporates methods for correct bit transmission. It adds redundant bits for error correction through convolutional coding and spreads data transmission by interleaving. The second GSM layer (Layer 2), referred to as the "data link layer", consists of an intelligent entity responsible for the secure transmission of data messages between the mobile and the base station. To do this, the transmit side structures the data messages from the higher layer to match the physical constraints of the Layer 1 medium. It requests an acknowledgement of the sent data from the receiver so that any packet that was not received can be sent again. At the receive side of Layer 2, messages are reconstructed from the received frames and the acknowledgement is formulated. The third GSM layer (Layer 3), also referred to as the "network layer", is responsible for the management of an established connection and of the associated activities in the radio network. In GSM, these tasks are further subdivided into the following sublayers: - Call control management CC - Mobility management MM - Radio resource management RR Rohde & Schwarz Training Center V 1.0

Fundamentals: Signalling at the Air-Interface

Page 3

Implementation of OSI Layers 1 to 3 at the GSM Air Interface:

Speech

Speech User data

User data Signalling

Frame building Acknowledge request

O S I L ayer 3 n e tw o r k

O S I L ayer 2 D a ta -T ra n s fe r

Channel coding error correction Interleaving

Signalling

Frame concatenation Acknowledgement

Error detection Deinterleaving

Equalisation RFM o d u la tio n

O S I L ayer 1 P h y s ic a l L a y e r

T r a n s m itte r

RFD e m o d u la tio n

R e c e iv e r

Fig. 8: Block diagram of GSM mobile station

As can be seen from the greatly simplified functional block diagram of the GSM transmitter and receiver, this segregation of functions not only provides a useful basis for apportioning the design effort, but also ensures that the measurement interfaces within the system are clearly defined.

Rohde & Schwarz Training Center V 1.0

Fundamentals: Signalling at the Air-Interface

Page 4

Implementation of Layer 1 at the GSM Air Interface: As has been said already, Layer 1 of the OSI reference model in GSM is the physical layer. A few aspects have already been described in the section dealing with the air interface: the burst types required for data transmission, the physical and logical channels used and GMSK (Gaussian minimum shift keying). A few other features required to support data transmission are described in the following. Channel Coding in GSM [GSM05.03] As the propagation conditions in mobile radio channels are highly variable, bit error rates (BER) as high as 10-3 to 10-1 can occur. To obtain high-quality and highly compressed speech and data communication, the BER has to be reduced to a rate of 10-5 to 10-6 through the use of appropriate error correction methods. This is the purpose of GSM channel coding. GSM channel coding is a combination of several methods each appropriately modified for use in the various logical channels. This section gives an overview of these methods and of derived components, the BCCH being used as an example to illustrate the special forms of application.

Components of GSM Channel Coding GSM channel coding comprises block coding for error detection, convolutional coding for error correction and interleaving to eliminate error bursts. Channel coding is performed in this order at the data source and in the reverse order at the data sink. The Fig. below gives an overview of channel coding and its components. The terms "outer" and "inner" error control indicate the position of the components in the transmission chain. The following sections describe each of the components and outline their hardware implementations. Outer error control Inner error control

Block coding

Convolutional coding

Interleaving

Fig.: GSM channel coding components

Rohde & Schwarz Training Center V 1.0

Deinterleaving

Convolutional decoding

Parity check

Fundamentals: Signalling at the Air-Interface

Page 5

Block Coding Block coding involves calculating a certain number of parity bits from a data-bit block and then appending them to the data block. At the data sink, i.e. the receive side, errors in the received code word can be detected with these parity bits. GSM uses two types of block code - CRC, (cyclic redundancy check) and a FIRE code (short, cyclic, binary code). The cyclic codes of this kind are also called (n,k) codes, n representing the number of code symbols (bits) and k the number of data symbols (bits). The number of check bits is therefore n-k. These check bits, and so the code word, are produced by a generating polynomial. A polynomial is used to represent words, the power of each term in the polynomial corresponding to a bit position and the coefficient of each term to a bit (Dm). A data word with k bits is, therefore, represented as: Dk ⋅ x k + Dk −1 ⋅ x k −1 + ... + D1 ⋅ x + D0 .

To calculate the check bits, the data word D(x) is multiplied by xn-k and then divided by the generating polynomial G(x) which is of degree (n-k). The remainder R(x) is the check word comprising the check bits:

 x n −k ⋅ D( x )  R( x ) = Re st  .  G( x)  The code word (C(x) is now obtained by appending the check word to the data word. C ( x ) = x n −k ⋅ D ( x ) + R ( x ) If there are no transmission errors, the code word C(x) is divisible by G(x). The probability is, therefore, high that errors will be detected. The division (and multiplication) of polynomials is easiest to implement with a feedback shift register. The type of feedback is determined by the generating polynomial G(x), the arithmetical operation (multiplication/division) by the feedback direction. A combination of multiplication and division by a shift register is also possible. A shift register of this type is shown in the Fig. below. The significant feature is that the data word D(x) is first multiplied by xn-k and then divided by the generating polynomial G(x) of degree n-k.

Input Fig.: Coder performing automatic multiplication by xn-k Rohde & Schwarz Training Center V 1.0

Output

Fundamentals: Signalling at the Air-Interface

Page 6

When the shift register performs division, the remainder left by division remains in the register after the dividend has been completely shifted through. If this remainder, i.e. the check word, is appended to the shifted data word and the code word obtained in this way is shifted through an identical shift register, only 0s will remain in the register because in the case of an error-free transmission a code word without errors must be divisible without a remainder. The check word is appended by the shift register shown in the Fig. above. The data word is not only shifted into the register but also applied to the output. The gate is blocked after k clocks and the register content is shifted to the output. The block codes could also be used for correcting the detected errors but this capability is not used by GSM.

Forward Error Correction, FEC, or Convolutional Coding: With FEC (forward error correction), redundant bits are inserted into data packets (bursts) at the transmit end to ensure that most of the information is not lost if there is a burst error. A convolutional coder is used. A convolutional coder "remembers" the last n bits sent and adds each input bit to the stored n bits. The words obtained at the output are usually longer than 1 bit. Error correction is based on the fact that a previous state, i.e. a word, or a bit sequence, can only assume one of two new states depending on whether 0 or 1 has been entered. If a word arrives at the receiver in a state that cannot be reached from a state obtainable from one of the two input combinations, a transmission error has occurred and needs to be corrected. This procedure is equivalent to tracing a path through a trellis diagram which is familiar from coding theory. Convolutional coding in GSM supports error correction by generating the necessary redundancy. The amount of redundancy is determined by the ratio of the data block length to the code block length, i.e. the coding rate r, of the convolutional coder. Like cyclic block codes, convolutional codes, too, can be represented by polynomials and realized by means of shift registers. The shift register has k memory locations which are read as defined by ν generating polynomials. The Fig. below shows a shift register with the constraint length K=5 used in GSM and the ν=2 generating polynomials. GSM defines seven different generating polynomials G0 to G6 for convolutional coding (DSM05.03, Ann. B). In the Fig. below, the following polynomials are used: G0 = 1 + D 3 + D 4 G1 = 1 + D + D 3 + D 4

Fig.: Block diagram of a convolutional coder Rohde & Schwarz Training Center V 1.0

Fundamentals: Signalling at the Air-Interface

Page 7

After the data bits have been clocked into the register, the outputs of the flip-flops corresponding to each of the generating polynomials are added mod 2. Each data bit is, therefore, XORed with (K-1) preceding bits; the convolutional coder is also said to have a "memory" of (K-1) bits. Each data bit is, therefore mapped onto ν=2 code bits with each clock cycle. The convolutional coder has a rate of r = 1/ν = 1/2 in this case. While K=5 is selected for all channels in GSM, the rate r varies depending on the number of generating polynomials used (GSM05.03, Sec. 3 and 4). The Figure above also shows why the term "convolution" is used to describe this code – something that is hardly ever made clear. Each generating polynomial and the data word are multiplied together each time the data word is shifted and the results are added. This is the mathematical operation of "convolution". Error correction is performed during decoding, generally using the Viterbi algorithm. The ability to correct errors increases with increasing K and decreasing r, but a small r, i.e. high redundancy, reduces transmission rates.

Interleaving Interleaving means that the information to be transmitted is spread or distributed over several bursts in such a way that contiguous information is split up and transmitted in several bursts. The original bursts can be regenerated with the aid of FEC even if one of the bursts is lost. For instance, the information in 4 bursts is divided into 2 x 4 blocks, one block containing the even bits and the other the odd bits. Two even blocks or two odd block are then always combined to form a burst. The result obtained when the convolutional code is decoded strongly depends on the frequency of occurrence and the distribution of bit errors. Extended periods of fading in the mobile channel cause successive bit errors, referred to as "error bursts". These error bursts have a particularly adverse effect on decoding. To avoid error bursts, an attempt is made to spread the bit errors over several code words. This is achieved by interleaving several code words. This method is also called diagonal interleaving (see Fig. below). Another kind of interleaving is block interleaving. Blocks of code words are written row-by-row into a matrix and then read column-by-column as shown in the Fig. below. With both methods, consecutive bits of a code word are never transmitted consecutively, and conversely, when the bits are de-interleaved at the receive end, error bursts are spread over several code words. As several code words are interleaved, the decoder has to "wait" a certain time until all bits of a particular code word arrive. This delay, i.e. the "measure for spreading over time" is referred to as the "interleaving depth". The greater the interleaving depth, the more code words are available for spreading the error bursts and the greater the probability that errored bits can be corrected. However, the greater the interleaving depth, the greater the transmission delay. The different types of interleaving and interleaving rules used for the various channels are specified by GSM05.03, sec 3 and 4).

Rohde & Schwarz Training Center V 1.0

Fundamentals: Signalling at the Air-Interface

Spreading

Interleaving

Fig.: Diagonal interleaving

Write

Read

Fig.: Block interleaving

Rohde & Schwarz Training Center V 1.0

Page 8

Fundamentals: Signalling at the Air-Interface

Speech frame

Page 9

Signalling

260 Bits

184Bits

Class 1 cyclic code 267 Bits

FIRE code 229 Bits

Convolutional Coding 456 Bits

Reorganization, Segregation Insertion of Stealing Flags 456 Bits in 8 subblocks

Block diagonal interleave

Block rectangular interleave

Intra-Burst interleave

Ciphering

To Modulator

Fig.: Error correction coding at Layer 1 of the GSM air interface Rohde & Schwarz Training Center V 1.0

Fundamentals: Signalling at the Air-Interface

Page 10

Implementation of Layer 2 at the GSM air interface: The GSM Layer 2 is the data transmission layer, i.e. it is responsible for a reliable transmission of the data packets of Layer 3. To perform this task, it uses the physical transmission services of Layer 1. The LAPDm protocol, the link access protocol for channel D, is used in this case. This protocol used for the data transmission layer is familiar from ISDN, but is used in a slightly modified form at the GSM air interface. The "m" in the protocol designation stands for "modified". The reason for this modification is the low data transmission rate at the air interface. Signaling blocks have, therefore, been omitted whenever possible. In contrast to the LAPD protocol used for ISDN, the LAPDm protocol does not contain start and end flags for synchronization or a terminal end point identifier for the terminal equipment in question. Since point-to-point is used at the air interface, terminal identification is not required. The LAPDm protocol does not require an FCS (frame check sequence), used for identifying transmission errors, either. In GSM, error control is performed by the physical layer. LAPDm Frame Formats Different frames are used by the LAPDm protocol for transmitting different kinds of data, either data associated with higher layers or Layer 2 signalling data. Three frame formats are defined. A length of 23 octets is specified for the frames, i.e. each frame consists of 23*8 = 184 bits. We can now see how this relates to Layer 1, where a sequence of max. 184 bits is specified for signalling before channel coding. The general frame format is as follows:

1octet

Adress Field

1octet

Control Field

1octet

Frame Length

N201 octets

Information Field

Fig.: General frame format of the LAPDm protocol The individual fields can be further subdivided. The following structures are obtained.

Rohde & Schwarz Training Center V 1.0

Fundamentals: Signalling at the Air-Interface

Page 11

Address Field: Structure of address field: Bit

8

7

6

5

LPD

4

3

SAPI

2

1

C/R

EA 1

EA bit: This bit marks the end or the beginning of the address field. If the bit is 0, another address octet will follow. If the bit is 1, this is the last octet of the address field. Remember that an LAPDm address field is always 1 octet long. This means that the value of this bit at the GSM air interface is always 1.

C/R bit, command/response bit: This bit indicates whether the present frame is a command or a response and the initiator of the command. The following values indicate the direction:

Command: C/R = 0 Response: C/R = 0

Command: C/R = 1 Response: C/R = 1

Link protocol discriminator LPD: The GSM link protocol discriminator always has the binary code 00 with a single exception: messages of the cell broadcast channel CBCH are assigned the LPD code 01.

Rohde & Schwarz Training Center V 1.0

Fundamentals: Signalling at the Air-Interface

Page 12

Service access point identifier SAPI: The service access point is the interface between two protocol layers. It defines the "handover point" where data to be transmitted are received, or received data handed on to Layer 3. In GSM, the following two decimal values are currently defined for SAPI at the air interface. Decimal value of SAPI 0

Explanation Radio resource RR Mobility management MM Call control CC Short message service SMS Supplementary services SS

3

Control Field: The Layer 2 control field specifies the frame type and contains the transmit and receive sequence number for I-frames. In the LAPDm protocol, the length of the control field is always limited to one octet. No general structure can be shown for the control field as the structure is different for different frame types. The frame types and control field structures defined by GSM are described in detail in the sequel.

Frame Length Field The frame length specifies the total length of the information field. Values between 0 and N201 are permissible. Structure of frame length field: Bit

8

7

6

5

Length

4

3

2

1

M

EL

EL bit: This bit indicates whether this octet is the last one in the frame length field or not. EL = 1 indicates the end of the frame length field. With GSM, a frame for the air interface is always 1 octet long, i.e. this bit is always set to 1. There is an option for longer frame formats in the future. M bit: If a message is longer than one frame, the M bit indicates that another information frame will follow. In this case M = 1. If M is 0, the frame is the last frame in the message.

Rohde & Schwarz Training Center V 1.0

Fundamentals: Signalling at the Air-Interface

Page 13

Frame Formats for Layer 2 of the GSM Air Interface: The general frame format is sub-classified using 3 further frame types: Bbis format, B format and A format. What distinguishes the frames are the logic channels they are used for and the type of information they transmit. The following frames are defined:

A Frame: The A frame can be used on all DCCHs (dedicated control channels) on the uplink and the downlink. An A frame is sent when it is not necessary to send signalling information after a connection has been established. This means that the A frame is a filler frame used as pseudo signalling. The structure of the A frame, particularly that of the control field, is identical to that of the B frame, but the information field in frame A contains only filler bits instead of Layer 3 signalling information. This frame format is sent, say, directly after channel setup, provided it is not necessary to send Layer 3 signalling information.

B Frame: The B frame is used for "real" signalling. Its information field contains the Layer 3 information. The B frame is used in all signalling channels, ACCHs (associated control channels) and DCCHs (dedicated control channels). The constant N201 defines the length of the information field, which is different for the different channel types although the whole frame is still 23 octets long.

Structure of the A Frame and B Frame Control Field for Different Frame Types: As mentioned already, the control field distinguishes between the frame types used for the different applications. As shown in the Fig. below, several frame types are used.

P, F and P/F bit, polling and final bit: This bit is contained in all control frames. It is called a polling bit in command frames and a final bit in response frames. In frames which may be used for a command or a response, either the polling bit or the final bit can be used. This bit indicates that the transmit side expects a response from the receive side although this is not compulsory for the frame type. P bit = 1 means that the transmitter of the command frame is waiting for a response from the receiver. The response must be given as F bit = 1.

Rohde & Schwarz Training Center V 1.0

Fundamentals: Signalling at the Air-Interface

Page 14 Information-Frame

Bit

7

6

5

N(R)

4

3

P

2 N(S)

1 2

0 0

I-Frame Supervisory-Frames

N(R)

P/F

0

0

0

1

N(R)

P/F

0

1

0

1

N(R)

P/F

1

0

0

1

RR-Frame

RNR-Frame

REJ-Frame Unnumbered-Frames

0

1

1

P

1

1

1

1

SABM-Frame

0

0

0

F

1

1

1

1

DM-Frame

0

0

0

P

0

0

1

1

UI-Frame

0

1

0

P

0

0

1

1

DISC-Frame

0

1

1

F

0

0

1

1

UA-Frame

Fig.: Structure of A and B frame control field, different frame types Information Frame, I-frame: The information frame is used to transmit Layer 3 signalling data. It is the transport frame for this kind of information. Each control field contains a 3-bit counter for the sent frames N(S) and the received frames (N(R). The N(S) counter indicates the number of frames marked by the transmitter as sent and the N(R) counter the number of frames the receiver indicates as received, i.e. this counter practically acts as a confirmation field for correct information transmission. Note that, due to the 3-bit coding of the N(R) count, not more than 8 I-frames can be sent without a confirmation. Rohde & Schwarz Training Center V 1.0

Fundamentals: Signalling at the Air-Interface

Page 15

Supervisory Frames: The supervisory frames group contains three different types of frame: RR, RNR and REJ. These frames are used for Layer 2 peer-to-peer signalling. Receive ready frame RR: RR frames are used to acknowledge I-frames. They contain the N(R) field, i.e. the receiver tells the transmitter the number of the last frame which has been confirmed as received. If the transmitter has sent more frames than the receiver confirms, the transmitter knows that all the frames after this number are not confirmed and therefore have to be sent again. Polling is also performed with the RR frame. This means that each of two communicating parties checks at regular intervals whether the other can still be reached (i.e. mutual polling).

Receive not ready frame RNR: With this message, the receiver indicates that it cannot receive more I-frames at present. This means that the transmitter has to stop sending I-frames until a higher layer signals a call clear-down or until data transmission can be continued.

Reject frame REJ: Unlike RNR which signals an overload, this message tells the transmit end that the received I-frame was faulty and has been rejected. The counter N(R) again indicates the number of the frame from which data must be retransmitted.

Unnumbered Frames: The unnumbered frames are another group of signalling frames used peer-to-peer at Layer 2. From the structure of the control field we see that the field does not contain a sequence number, i.e. that it does not confirm frames, and so is referred to as an unnumbered frame. The following frame types are available at the GSM air interface. Set asynchronous balanced mode frame SABM: This frame initiates a Layer 2 connection, i.e. one end requests the other to change to the "balanced mode" where both ends can exchange information on a peer-to-peer basis. This frame can be thought of as setting up a peer-to-peer connection and is active until a Layer 2 connection has been made. Disconnected mode frame DM: This frame indicates that the Layer 2 link cannot be maintained. The transmit end then indicates that it will immediately terminate the connection and not wait for any confirmation from the receive end. The DM frame is one way of terminating a Layer 2 connection.

Unnumbered information frame UI: This is another frame for transmitting Layer 3 information. However, unlike the normally used I-frame, this frame does not contain a transmit and receive sequence number. Rohde & Schwarz Training Center V 1.0

Fundamentals: Signalling at the Air-Interface

Page 16

The UI frame is, therefore, used when Layer 3 data are not referenced and transmitted in the non-acknowledge mode. The receive end need not send an RR message to confirm reception of an UI frame.

Disconnect frame DISC: This frame too terminates a Layer 2 connection. The transmit end uses the DISC frame to inform the receive end that it intends to terminate the Layer 2 connection. The transmitter however waits until it receives a UA frame as acknowledgement.

Unnumbered acknowledgement frame UA: The UA frame is the receive end’s response to an SABM or DISC frame. It is used to confirm the successful setup of a Layer 2 connection and also its termination.

Bbis Frame: Unlike the other two frames, the Bbis format does not use a header. The reason for this is that it is only used on the broadcast control channel BCCH and the common control channels AGCH (access grant channel) and PCH (paging channel) and that it is for all intents and purposes generally valid. This makes detailed addressing of a specific receiver with a protocol header superfluous. CCCHs (common control channels) use the point-to-multipoint mode for information transmission. This, therefore, simplifies the general structure of a Bbis frame:

N201 octets (here N201 = 23)

Information Field (Layer 3 Messages)

Fig.: Structure of Bbis frame

Rohde & Schwarz Training Center V 1.0

Fundamentals: Signalling at the Air-Interface

Page 17

Implementing Layer 3 at the GSM Air Interface: As described already, Layer 3 data are embedded in the protocol information of Layer 2. The general structure of a layer 3 message comprises three fields: type identifier, message type and data field (see Figure below). Layer 2 Header 1octet

Typ Identifier

1octet

Message Type

Data Field

Fig.: Structure of a Layer 3 message Type Identifier: The 8-bit type identifier contains the protocol discriminator which divides the information into different groups. Structure of the type identifier: Typ Identifier

CC-Messages

TI-Value

TI-Flag

MM / RRMessages

Skip Indicator „0000“

Protocol Discriminator

Protocol Discriminator

Fig.: Structure of type identifier

Protocol Discriminator PD: The 4-bit protocol discriminator divides the Layer 3 messages into different groups permitting different users to be addressed within this layer. Each message is assigned to one and only one of the groups. With the aid of the PD the following message groups are distinguished: Protocol discriminator 06 05 03

User group Radio resource management (RR) Mobility management (MM) Call control (CC) Supplementary services (SS) Short message service (SMS)

Rohde & Schwarz Training Center V 1.0

Fundamentals: Signalling at the Air-Interface

Page 18

The three most important groups in Layer 3 are the RR, MM and CC functions: Radio Resource Management RR: The radio resource management is responsible for the management and organization of radio connections at the air interface. This includes, for instance, the organization of physical and logical channels at the air interface and functions for setting up and maintaining a radio link. In contrast to the two other message groups, RR management in the network is mainly performed by the base station subsystem BSS, i.e. by the base station (BTS) and the base station controller (BSC). Some messages are, however, evaluated by the MSC (mobile switching center). The message type decides which messages these are. It then provides a "tunnel" through the BSS to the MSC. Mobility Management MM: The main task of mobility management is to obtain and store information about the location of subscribers in the network. Examples are the identification and evaluation of the cell identity at the mobile end or the storage of the location area in the VLR (visitor location register). The MM uses the channels provided by the radio resource for transparent data transmission between the mobile station and the MSC. An example is the location update. First, a radio link is established, then data are sent from the MS to the MSC. The data are acknowledged and the radio link is terminated. This illustrates the hierarchy within the 3 message groups: MM is an application layer for the radio source management. Call Control CC: With the aid of the call control function, connections are set up and maintained. This not only applies to the air interface; a connection is managed through to the ISDN terminal at the subscriber end. Call control also uses the services of the RR radio resource management. Skip Indicator: The protocol discriminator of an MM or RR message is followed by the 4-bit skip indicator. This indicator has no specific function in GSM at present and is permanently set to 0000. Any other coding should be ignored at the receiver end. Transaction Identifier TI: If the Layer 3 message is from the call control management group, the skip indicator is replaced by the transaction identifier TI. It consists of the TI flag and the TI value. Before an explanation is given, an example will illustrate why the TI is needed: With call control, it is possible for a user to set up two or more transactions. Let us assume a mobile originated call (MoC) is in progress. The user goes to hold and calls another subscriber. The CC function now has to manage a second connection. The transaction identifier makes it possible to distinguish between several simultaneous transactions. The TI flag differentiates between the side initiating the call (TI flag = 0) and the side that is responding to the transaction (TI flag = 1). In the case of a mobile originated call, the TI flag of the call control message is, therefore, always set to 0. The TI flag for the response messages from the NSS (network subsystem) is always set to 1. The call-initiating end also assigns a TI value between 0 and 6. Rohde & Schwarz Training Center V 1.0

Fundamentals: Signalling at the Air-Interface

Page 19

Message Type: The message type indicates what the message is for, i.e. a 1-octet designator contains the message name and the command. GSM specification 04.08 defines all available message types, briefly describes what they are for and specifies the parameters they contain. The further structure of the information or data field depends on the message type. Each message is normally assigned a compulsory data field consisting of mandatory information elements IEs and another data field containing optional information elements. Both fields can have a fixed or variable length, depending on the message type. If the parameters following the IEI (information element identifier) field do not have a fixed length, the actual length of the parameter field is indicated in a length field.

Rohde & Schwarz Training Center V 1.0