Tài liệu Điện thoại di động mạng lưới Radio P9 pdf

The home station con-sists of a DECT fixed system DFS, which controls the system, a simpledatabase DB for user administration and a fixed part FP, which pro-vides the radio supply to the

Besides cellular mobile radio networks that are primarily envisaged for useoutdoors, systems that have been specifically designed for use in buildingsare also important In recent years cordless telephones with a range of afew hundred metres have become increasingly popular in private households.There is (along with CT2/CAI) a digital alternative to these analogue devicesthat offers better voice quality and a greater security against eavesdropping,

as well as other advantages: the DECT system

The abbreviation DECT originally stood for Digital European CordlessTelecommunications, but to underline its claim of being a worldwide stan-dard for cordless telephony DECT today stands for Digital Enhanced CordlessTelecommunications This standard was specified by the European Telecom-munications Standards Institute (ETSI) in 1992 A DECT network is a mi-crocellular digital mobile radio network for high user density and primarilyfor use inside buildings However, an outdoor application is also possible.DECT systems allow complete cordless private branch exchanges to be set

up in office buildings Calls can be made over the normal subscriber line aswell as between mobile stations through the DECT base station subsystem.When a staff member leaves his office, he usually can no longer be reachedfor incoming calls over a wireline telephone connection, although he may only

be located in another part of the building But if he uses a DECT terminal,

he can continue to be reached under his normal telephone number wherever

he may be anywhere on the premises To enable users to continue to receivecalls after they have left the DECT coverage area, ETSI specified an interfacebetween GSM and DECT (see Section 9.16)

The DECT standard permits the transmission of voice and data signals.Consequently cordless data networks can also be set up on a DECT basis.The use of ISDN services (Integrated Services Digital Network ) is also pos-sible Users are able to move freely within different cells without risking aninterruption to their calls The handover process switches calls from one radiocell to the next without interrupting the call

In outdoor areas the maximum distance between base and mobile station isapproximately 300 m; in buildings, depending on the location, it is up to 50 m.Larger distances to the base station can be bridged through the installation

of appropriate base stations using the relay concept (see also Section 9.12)

∗ With the collaboration of Christian Plenge and Markus Scheibenbogen

The first time DECT systems were presented to the public was at theCeBIT exhibition in Hanover in March 1993.

Since then, the costs for a DECT mobile have been decreased to be parable to those of analogue cordless telephones

The size of a DECT system determines how it is installed The labels large andsmall are relative and relate to the number of mobile stations to be operatedwithin the DECT coverage area A DECT system is capable of automaticallylocalizing up to 1000 subscribers in one location area (LA) With a largernumber of users additional location areas are required that fall under theadministration of the DECT system (see Sections 9.1.2 and 9.16.1.3)

DECT fixed networks are normally designed as dedicated private branch changes

ex-If the DECT systems are installed by a network operator, each specific DECT system can be allocated its own location area All DECTlocation areas are then interconnected over a backbone ring and administeredcentrally by a DECT system, as illustrated in Figure 9.8 Each customer hashis own DECT fixed station (DFS) with his own identification of the locationarea

customer-Despite the efficient location administration of DECT systems, it wasplanned that each customer would be allocated his own location area to en-sure that the channel capacity of other customers would not be affected bythe calling activity of his own mobile stations

Private home base stations These offer a possibility of using the DECT tem in small private households (see Figure 9.1) The home base stationsupplies the entire area of a house and can support one or several mobilestations that are supplied by the base station The home station con-sists of a DECT fixed system (DFS), which controls the system, a simpledatabase (DB) for user administration and a fixed part (FP), which pro-vides the radio supply to the mobile stations An interworking unit(IWU) is provided for connection to an external network [20]

sys-Wireless private branch exchanges In terms of installation, central systemswith a DFS to which several FPs are connected are suitable for largeprivate households or small companies (see Figure 9.2) Fixed terminalscan also be connected The system is administered by the DECT fixedsystem The correct location area of the mobile stations is stored in

a database (DB), and an IWU enables a connection to the externalnetworks

ISDN PSTN GSM

IWU DB

= Portable Part PP

Figure 9.1: Private home base station

IWU

DB

ISDN PSTN GSM

FP

FP PP

PP

DFS

Figure 9.2: Wireless private branch exchanges

Public Telepoint systems These systems provide DECT mobile stations cess to the public telephone network over “other” FPs at public sites (seeFigure 9.3) Possible application areas are public facilities with high userdensity such as airports, train stations and city centres These privateand public DECT systems consist of a number of FPs that are admin-istered by a DECT fixed system (DFS) Access to the public telephonenetwork can be provided by a network interface (With DECT, PPs inprinciple can also be reached in the area of a Telepoint The restriction

ac-to outgoing calls is dictated by the respective licensing agreement).Wireless local rings This is a ring to which a terminal adaptor (TA) and allthe terminals are linked (see Figure 9.4) The terminal adaptor pro-vides a connection over the radio interface to one of the public FPs inthe area The terminals (e.g., telephone, facsimile) are linked to the ter-minal adaptor over conventional lines If necessary, the terminal adaptor

ISDN PSTN GSM PP

PP

FP

FP Airport

Railway Station

FP City Centre

IWU

DB DFS

Figure 9.3: Public Telepoint system

TA

ISDN PSTN GSM IWU

TA = Terminal Adapter

FP

DB DFS

Figure 9.4: Wireless local rings

IWU

ISDN PSTN GSM DB

FP PP

PP

DFS

Figure 9.5: Neighbourhood Telepoint

establishes a radio connection to the next FP The FP is administered

in the same way as with the public Telepoint service

Neighbourhood Telepoint This is a combination of the public Telepoint vice and a private home base station (see Figure 9.5) Private households

ser-do not have their own FP; instead one FP supplies a number of ent private households on the basis of the Telepoint principle The FPs

SCU

SCU

Backbone Ring

ISDN PSTN GSM PP

Figure 9.6: Private branch exchanges with ring and central DFS

are installed so that several households will always be supplied by one

FP A DECT fixed system connected over a network interface with aninterworking unit (IWU) to the public telephone network manages theFPs

Because of the way they are designed in terms of size and subscriber ity, the following systems are also suitable for use by companies with scatteredoperating locations One or more nodes can be provided for DECT at eachcompany location Each implementation of the system must be flexible andtailored to the needs of the customer in each individual case

capac-Private branch exchanges with ring and central DFS These private branchexchanges consist of a backbone ring with a number of switching nodes(see Figure 9.6) Each switching node consists of a Subsystem ControlUnit (SCU) and several base stations, i.e., FPs One of the nodes con-tains the DFS which is responsible for all the nodes connected to thebackbone ring and controls them from a central function The DFS alsocontains the home database (HDB) and the interworking unit (IWU),which constitutes the external connection All signalling data transmit-ted in this system must be transmitted over several nodes (SCUs) aswell as the backbone ring Transmission channels of the network thatare no longer available for transporting user data are used for this pur-pose A disadvantage is that the entire network suffers if the centralsystem breaks down Some relief is provided by the decentral systems,which ensure that there is a reduction in signalling traffic and that only

IWU

SCU

Backbone Ring

IWU

ISDN PSTN GSM

VDB HDB

VDB HDB

HDB = Home Data Base VDB = Visitor Data Base

DFS

DFS PP

PP

FP FP

FP

FP PP

PP

Figure 9.7: Private branch exchanges with ring and distributed DFS

the respective mobile stations are affected when the subsystem breaksdown [20]

Private branch exchanges with ring and decentral DFS These are set upsimilarly to the systems with a central DFS (see Figure 9.7) and consist

of a backbone ring that interconnects a number of nodes A control unit

is located at each of these nodes and several fixed stations are connected

to a node

This system operates as a decentral structure because each control unit

is controlled by its own DECT fixed system Furthermore, each DECTfixed system has an interworking unit (IWU) that connects it to theexternal networks The data is also stored at decentral locations Sub-scriber administration is carried out in a home database (HDB), fromwhich important data for operations can be retrieved over the network.For the purposes of reducing the signalling required when a mobile sub-scriber is roaming the network, certain information about each sub-scriber who moves out of the coverage area of his home database iswritten to the visitor database (VDB) of the new coverage area Thenetwork provides access to the data in the visitor database, and thebackbone ring is unloaded of responsibility for the signalling If a DFSsuffers a breakdown, only the subscribers moving in the respective cov-erage area will be affected [20]

Private branch exchanges with direct connection of FP These branch changes have the same decentral structure as the decentral private

ISDN PSTN GSM

VDB HDB

VDB HDB

Figure 9.8: Private branch exchanges with direct connection of FP

branch exchanges (see Figure 9.8) The difference between them is thatthe fixed stations of these private branch exchanges are directly con-nected to the respective DFS, which has the advantage that there is

a reduction in the signalling traffic within a node and consequently ahigher capacity for user data [20]

a home location register (HLR) and a visitor location register (VLR).Incoming calls are routed over the HLR to the VLR and from there tothe mobile terminal Outgoing calls only use the VLR Access of theHLR is not required for these calls

X.500 Another method is the storage of data based on the standardized ister service ITU-T X.500 The database is distributed physically overdifferent locations but logically centralized The data therefore givesthe impression of being stored in a Directory Information Base (DIB).All data is organized as objects in a hierarchical Directory InformationTree (DIT), in which each branch can be filed in a physically separateDirectory System Agent A Directory User Agent accesses the individual

reg-objects The search for the data takes place in a chain of interlinkeddatabases A critical factor of this method is the time involved in access-ing the data, but this is less of a problem for private branch exchangeswith a limit on subscriber numbers.

Telecommunication Management Network (TMN) A further method fordata storage is TMN based on ITU-T M.30 A management system

is controlled by a “manager” that realizes the management system usingobjects containing important data These objects are stored in a hierar-chical MIB database (management information base) This MIB can bedistributed physically over different locations The management system

is able to reproduce the HLR and VLR databases and store all locationarea data in one object

In comparison with public networks, only a limited number of users aremanaged in private networks such as DECT branch exchanges The data forthese users can usually be administered in one register Moreover, it is easier

to change objects in the frequently changing organizational structure of officeenvironments than in conventional databases This makes the HLR/VLRconcept less suitable for private branch exchanges than the other alternatives

A comparison between X.500 and TMN shows that X.500 is more flexiblebecause it is not required to copy the HLR/VLR principle

The DECT Reference System describes the defined logical and physical ponents of the DECT system, the interfaces between the different units andthe connection points to other networks The global logical structuring of

com-a loccom-al DECT network is explcom-ained below This is followed by excom-amples ofdifferent physical implementations

The logical groups of the DECT network are organized according to theirfunctionalities with the intervening interfaces D1, D2, D3 and D4, which,however, do not describe how they are physically implemented (see Figure 9.9)

9.2.1.1 Global Network

The global network supports the national telecommunication services Itcarries out address conversion, routing and relaying between the individualconnected socalled local networks The global network is usually a national(sometimes an international) network Examples of local networks include:

• Public Switched Telephone Network (PSTN)

Portable Application PortableApplication

Portable Application Portable Application

Portable Radio Termin.

Portable Radio Termin.

Portable Radio Termin.

Portable Radio Termin.

Portable Radio Termin.

Fixed Radio Termination

Fixed Radio Termination

Figure 9.9: DECT Reference System: Logical grouping

• Integrated Services Digital Network (ISDN)

• Packet Switched Public Data Network (PSPDN)

• Public Land Mobile Network (PLMN)

9.2.1.2 Local Networks

Each local network provides a local telecommunications service Depending

on the actual installation, this can vary from a simple multiplexer to a developed complex network If the subordinate DECT fixed radio termination(FT) does not have a switching function then the local network must take overthis function Yet it should be noted that there can be a difference betweenlogical definition and physical implementation, e.g., it is possible for severalnetworks with their functions to be combined into one unit

highly-A local network converts the global identification numbers (e.g., ISDN bers) to the DECT-specific IPUI (International Portable User Identity) andTPUI (Temporary Portable User Identity ) The following networks are oftenfound under the local network:

num-• analogue or digital Private Automatic Branch Exchange (PABX)

• Integrated Services Private Branch Exchange (ISPBX)

• IEEE 802 LAN: local area network based on IEEE 802

All the typical network functions must be embedded outside the DECTsystem and occur either in a local or in a global network Similarly to GSM,

the HDB and VDB are required for controlling inter-DECT mobility, i.e., lowing users to use their mobile stations to move within different independentDECT areas (see Section 3.2.1.3) Incoming calls are automatically routed

al-to the subsystem in which the user is currently located When a user movesfrom one network to another, a new entry for the current VDB is made in theHDB

9.2.1.3 DECT Network

The DECT network consists of base and mobile stations, and connects users

to the local fixed network No application processes are defined for the system,and it only functions as a multiplexing facility A DECT system always hasonly one network address per user, i.e., mobile station, and (from a logicalstandpoint) consists of one or more fixed radio terminations (FT) and manyportable radio terminations (PT) that have been allocated to it

Fixed Radio Termination The FT is a logical grouping of all the functionsand procedures on the fixed network side of the DECT air interface It isresponsible for:

• layer-3 protocol processes in the C-(control) layer (except for mobility)

• layer-2 protocol processes in the U-(user) layer

• layer-2 switching (routing and relaying) in the respective DECT networkExcept for handover and multicell management, the FT contains no switch-ing functions Although it can manage a large number of call entities, it cannotestablish a direct connection between two users This can only be carried outoutside the logically delineated area of the FT in the local network

Portable Radio Termination and Portable Application These two parts stitute the logical groups on the mobile side of the DECT network Whereasthe portable radio termination with all its protocol elements for OSI layers 1, 2and 3 is defined in the standard, it is up to the manufacturer of the equipment

con-to define the acceptable application Therefore it is not standardized

Whereas the logical structure of the DECT network is clearly defined, thephysical grouping can assume different forms It is adapted to each cus-tomer’s requirements, and therefore can be conceptualized as a single basestation with up to 12 simultaneously communicating mobile stations if it isequipped with one transceiver or as an independent switching centre for officebuildings The logical interfaces D1, ,D4 are thereby partially integratedinto a common physical unit and consequently can no longer be clearly ap-portioned (see Figure 9.10)

RFP RFP

Figure 9.10: DECT Reference System: physical grouping

9.2.2.1 DECT Base Station (Fixed Part)

Physically a DECT system can be split into two parts: a DECT fixed part(FP) on the fixed side and a DECT portable part (PP) on the mobile side.The fixed part on the wireline side can contain one or more logical groups ofthe fixed radio termination with common control

An FP can be divided into two physical subgroups:

• Radio Fixed Part (RFP): responsible for only one cell in the network

• Radio End Point (REP): corresponds to a transceiver unit in the RFP

9.2.2.2 DECT Mobile Device (Portable Part)

The two logical groups portable radio termination and portable application arephysically combined into one portable part (PP), typically a handset A PPnormally only has one radio endpoint

A mobile station can be used by different users Each user must identifyhimself before being granted access to the DECT network This function ishandled by a DECT Authentication Module (DAM), which contains informa-tion for the identification (International Portable User Identity, IPUI) andauthentication (Authentication Key, K) of the user and can be inserted intothe PP The DAM contains all the required encryption procedures

The DECT system description [5] lists several typical DECT configurations.Different physical implementations are required, depending on the higher-ranking network:

Examples of possible installations are given below Several suppliers arealready marketing private household systems at reasonable prices Complex

private office installations are currently conquering the market In addition,DECT systems have been implemented and tested (since 1997) as Radio LocalLoop systems.

9.2.4.1 PSTN Reference Configuration

Domestic Telephone The private domestic telephone constitutes the plest DECT configuration Here the network is connected to a PSTN over asubscriber interface, just like a plain old telephone (POT) (see Figure 9.11).The functional characteristics resemble those of the earlier CT 1 and CT 2generations of cordless domestic telephones (see Chapter 8) There is no pro-vision for a local network

sim-PBX Similarly to the domestic telephone, in a simple implementation of aDECT-PBX, individual FPs are connected to the switching unit of the PSTN.Technically it would then be very complicated to allow a changeover from one

FP to another FP during a call (handover)

In Figure 9.12 the fixed part contains several radio fixed parts, each ofwhich always serves one cell, thereby giving the system its cellular character

A mobile station establishes a connection to the strongest RFP If the user

is roaming in the area of a neighbouring cell, the FP is then in a position

of executing an internal layer-2 handover The protocols for cell control arecontrolled by the common control (CC) function, which physically can beintegrated into the FP or into the PBX

Radio Local Loop A DECT system can also be incorporated as a local accessnetwork into the PSTN (see Figure 9.13) In this case the radio connection istransparent to the user The user’s wireline telephone is linked to a CordlessTerminal Adapter (CTA), which carries out the radio transmission to theRFP In the radio in the local loop (RLL) area, suppliers of fixed networksare currently trying to avoid the high cost of local network cabling by usingDECT-RLL systems for the so-called last mile A comparable configurationusing ISDN as the local fixed network and DECT as the RLL system is shown

in Figure 9.14

9.2.4.2 GSM Reference Configuration

Aside from the X.25 reference configuration, which is not covered here, theconnection between the two mobile systems GSM and DECT should be men-tioned This development offers users the possibility of coupling a locally basedDECT system with the national GSM mobile radio system (see Figure 9.15).From the standpoint of GSM, the portable application, PT, FT and possiblyalso a local network form a mobile user unit The D1 reference point is the

R reference point in the GSM standard (see Figure 3.3) Detailed coverage ofthe integration of DECT and GSM systems is provided in Section 9.16

Global

Network

Fixed Radio Term

Portable Radio Term

Portable Appl Local

Portable Radio Term

Portable Appl Local

Portable Radio Term

Portable Appl Local

Portable Radio Term

Portable Appl Local

ISDN Terminal ISPBX

U

Figure 9.14: Radio local loop configuration for ISDN

Global Network FixedRadio Term PortableRadio Term

Portable Appl Local

Figure 9.15: GSM-DECT configuration

The DECT reference model was designed in accordance with the ISO/OSImodel (see Section 2.5) Because the DECT system incorporates the interfacebetween communicating partners, only certain aspects of the applications-oriented layers are included in the standard, e.g., encryption

The key functions of the DECT system correspond to the three lower layers

of the ISO/OSI model: Physical, Data Link and Network layers Becausewith DECT the quality of the transmission medium (radio) is continuouslychanging and channel access is a complicated function that must be carriedout frequently, the data link layer has been divided into the two sublayers DataLink Control (DLC) and Medium Access Control (MAC) Figure 9.16 relatesthe DECT reference model to the corresponding ISO/OSI layers Above theMAC layer the functions of the layers are grouped into two parts: the Controlplane for signalling and the User plane for the transmission of user data Thecontrol functions of the C-plane are processed in the network layer, whereasthe data of the U-plane is passed on unprocessed

The DECT layers with their characteristics are briefly introduced below Adetailed description of the DECT layers is provided in Section 9.4

The physical layer (PHL) is responsible for implementing the transmissionchannels over the radio medium At the same time it has to share the mediumwith many other mobile stations that are also transmitting Interference andcollisions between communicating base and mobile stations are largely avoidedthanks to the decentrally organized use of the available dimensions location,time and frequency (see Figure 9.17) Each dimension contains several possi-bilities for seizing a channel for interference-free transmission

The TDMA (Time-Division Multiple-Access) method is the one used forthe time dimension Each station sets up its channel in any available time slot

Application Process Process

C-Plane

Interworking Signalling

Application

Layer

ControlNetwork

Physical LayerMedium-Access Control LayerLayer

Data Link

LayerControlData Link

Figure 9.16: DECT reference model

and is then able to transmit in that time slot at a constant bit rate Because

of the TDD (Time-Division Duplexing) method the uplink and downlink ofthis channel are in slot pairs on the same frequency Consequently a duplextransmission always occupies two time slots, which are separated from another

Key technical data for the DECT system is listed in Table 9.1 (seealso Table 8.2) With only 12 duplex voice channels in 1.73 MHz, i.e.,

144 kHz/channel pair, DECT is very generous in its use of spectrum (GSMonly requires 50 kHz/channel pair) However, because of its dynamic channelselection, the resulting small random reuse distances and the microcellularcoverage, DECT provides a much higher capacity (Erl./km²), producing al-most unbelievably high values (10 kErl./km²), especially in high-rise buildingsbecause of the reuse distances at every two levels over the same base area

Time Slot

BS 2

Figure 9.17: Three-dimensional use of spectrum

Table 9.1: Physical data on the DECT system

6.4 kbit/s signalling (A-field)

The MAC layer (Medium-Access Control (MAC) Layer ) (see Figure 9.16) isresponsible for setting up, operating and releasing channels (Bearer ) for thehigher layers The different data fields of the MAC protocol are protected bycyclical codes, which are used for error detection at the receiver The MAClayer ensures that service-specific control data is added to each time slot.The MAC layer comprises three groups of services:

BMC The Broadcast Message Control Service is offered on at least one ical channel in each cell, even if there is no user transmitting Thischannel is called a beacon channel, and produces a continuous connec-tionless point-to-multipoint connection on the downlink in which a basestation transmits its system-related data This identifies the base sta-tion to the mobile unit By evaluating the received signal, the terminalcan determine the current channel quality at the same time.

phys-CMC The Connectionless Message Control Service supports a less point-to-point or point-to-multipoint service that can be operatedbidirectionally between a base station and a mobile user

connection-MBC The Multi-Bearer Control Service offers a connection-oriented to-point service An entity transmitting in one or both directions is able

point-to support several bearers, thereby achieving a higher net data rate.Each of these three services incorporates its own independent service accesspoint (SAP), which links it to the next-highest layer and can combine severallogical channels together

The protocol stack directly above the MAC layer divides into two parallelparts Similarly to the MAC layer, extensive error protection in the C-plane

of the data link layer improves the reliability of the data transmission Alongwith a point-to-point service, the C-plane offers a broadcast service to thenetwork layer above it The U-plane is responsible for processing user data onthe radio section, with the spectrum of services ranging from the transmission

of unprotected data with minimal delays, e.g., speech, to the provision ofprotected services with variable delays for data transmission The requireddata rate of an existing connection can be changed at any time

The network layer sets up, manages and terminates connections between usersand the network The U-plane in DECT has no role in the network layer andforwards all data unprocessed in a vertical direction The C-plane carries outthe signalling and is responsible for controlling data exchange Five protocolsbased on the link control entity are provided for this purpose Along withthe call and connection entities, there is a mobility management service thatcarries out all the tasks necessary to support the mobility of mobile stations

In addition to the data for location area management, messages for cation and encryption data are also transmitted

authenti-9.3.6 Management of the Lower Layers

The management of layers 1–3 (Lower Layer Management Entity (LLME) (seeFigure 9.16) contains procedures that affect several protocol layers Processessuch as generating, maintaining and releasing physical channels (Bearers) areinitiated and controlled from this entity Furthermore, the LLME selectsfrom the physical channels which are available and evaluates the quality ofthe received signal

9.4.1.1 FDMA and Modulation Techniques

Because of FDMA access, the DECT system has the possibility of selectingfrom a number of frequencies in its channel selection

It operates in the 1880–1900 MHz frequency band Within this band 10carrier frequencies are defined, the mid-frequency fcof which can be calculated

Key-If a transmit signal is formed from two orthogonal bandpass signals with ferent mid-frequencies, this is referred to as Frequency Shift Keying (FSK) AGaussian filter acting as a low-pass filter (GFSK) removes the high-frequencyparts of the signal in order to keep the band of the signal’s spectrum as small

dif-as possible If the modulation index is 0.5 and the possibility of a coherentdemodulation of the radio signal exists, this shift procedure is referred to asminimum-shift keying (MSK) With GMSK the Gaussian filter also comesinto play For cost reasons, receivers in DECT systems are usually not builtwith coherent demodulation/modulation [27]

f479 f0




"!  #$

% &'( 

- -0/

/32 /54 /36 /57 /58 /39 /5.

//

Figure 9.18: Time-multiplex elements of the physical layer

The transmission of a binary 1 in a DECT system produces a frequencyincrease of 4f = 288 kHz to fc+ 288 kHz For the transmission of a 0 thefrequency is decreased by4f to fc− 288 kHz

The standard does not provide for any equalizers With a bit duration of0.9µs, waves that reach the receiver with a delay due to multipath propagation(see Figure 2.8) cause signal dispersion (see Section 2.1.7), which with a 300

m alternative path length already corresponds to the symbol duration andmakes reception impossible even when sufficient signalling power exists Theliterature recommends using 16 of the 32 synchronization bits in the S-field inaccordance with Figure 9.19 to estimate the impulse response of a channel inthe receiver, which equates to a good (but not one conforming to the standard)implementation of an equalizer [21]

9.4.1.2 TDMA Technique

Each station receives a protected, periodically recurring portion of the overalltransmission rate of a frequency According to Figure 9.18, the frame andtime slot structure of the DECT system is explained in the physical layer.The transmission capacity of each frequency is divided into 10 ms longperiodically recurring frames, each of which has a length corresponding tothe duration of 11.520 bits This produces a gross frame transmission rate of

1152 kbit/s A frame comprises 24 time slots, which are used as either fullslots, double slots or half slots (see Figure 9.18)

In a normal basic connection the first 12 time slots are used for transmittingdata from the base station to a mobile station (downlink ), whereas the secondpart of the 24 slots is reserved for the direction from the mobile station to thebase station (uplink ) Since a duplex connection requires both an uplink and a

Figure 9.19: The different physical packets in the DECT standard

downlink connection, the DECT system uses a technique called Time-DivisionDuplexing (TDD) If the base station occupies slot k in order to transmit tothe mobile unit, slot k + 12 is reserved for the mobile station to enable it tosend data to the base station This rigid system of allocation is abandonedwhen it comes to more complex advanced connections, where a generous use

of time slots is allowed in each transmission direction

Each of the 24 time slots has a length of 480 bits (416µs) and can be used

in accordance with the slot type (full, half, double) Different physical packetsare based on this structure Each physical packet contains a synchronizationfield S and a data field D A physical packet is shorter than a time slot byone guard period to prevent an overlapping of packets from neighbouring timeslots

A P00 (short physical packet ) is used for short connectionless transmissions

in the beacon channel or for short messages It comprises only 96 of the

480 bits and therefore provides a particularly large guard time in a slot This

is especially important at the start of a connection phase when a mobilestation is not yet completely synchronized with the network and could causeinterference to neighbouring time slots when it transmits a full slot

A P08j message only requires a half slot for transmission, and thereforebecause of the decreased transmission rate ends up with double the number

of available channels per frame; see Figure 9.19

Because full slots are usually the ones preferred when a slot is requested,the packet P32 will be described here The first 32 bits form the synchroniza-tion field S, which is used for clock and packet synchronization in the radionetwork It consists of a 16-bit preamble, followed by a 16-bit long packetsynchronization word In the S-field the response from the mobile stationcontains the inverted sequence of the bit of the synchronization field sent bythe base station

LockedIdle

Figure 9.20: Operating states of a mobile station

The S-field is followed by the user data field D, which is 388 bits long.Some of the content of this field is evaluated by the MAC layer, and thereforewill be covered in detail later (see Section 9.4.2.5)

The possibility of transmitting a so-called Z-field is offered by the D-field.This 4-bit long word contains a copy of the last 4 bits of the data field, which

is also referred to as an X-field A comparison of these two areas allows the ceiver to establish whether a transmission is being disrupted because of errors

re-in the synchronization of neighbourre-ing DECT systems These disruptions arereferred to as sliding collisions within the system A measurement of thesedisruptions results in the early detection of interference, and can be used asthe criterion for an optimal handover decision Packet P80 is the one with thehighest data rate, and requires a double slot User data rates up to 80 kbit/scan be achieved with this packet

9.4.2.1 Operating States of a Mobile Station

With reference to the MAC layer [9], a mobile station can find itself in one ofthe four states shown in Figure 9.20

• Active Locked The synchronized mobile station has at least one nection to one or more base stations

con-• Idle Locked The mobile station is synchronized with at least one basestation It currently does not have a connection, but is in a position toreceive requests for connections

• Active Unlocked The mobile station is not synchronized with any basestation, and therefore cannot receive any requests for connections Itattempts to find a suitable base station for synchronization to enable it

to move into the Idle Locked state

• Idle Unlocked The mobile station is not synchronized with any basestation, and is not in a position to detect base stations that are suitable

Active Traffic and Idle

Act Traffic

Active

Last released

First Traffic Channel established

and Dummy Bearer released

Dummy Bearer

established and Dummy Bearer released

Dummy Bearer released

establishedFigure 9.21: States of a base station

If a terminal is switched off, it is in the Idle Unlocked state When it isswitched on, a mobile station tries to transit to the Idle Locked state Itbegins to search for a suitable base station with which to synchronize If it

is successful, a transition to the Idle Locked state is made In this state amobile station is able to receive or transmit requests for connection As soon

as the first traffic channel is set up, it changes over to Active Locked state

If the last channel is cleared after the termination of a connection, the mobilestation returns to the Idle Locked state If it loses its synchronization withthe base station, it must change to the Active Unlocked state and seek anew suitable base station

9.4.2.2 Operating States of a Base Station

A base station can find itself in one of four states The Inactive state inwhich a base station is switched off is not shown in Figure 9.21

• Inactive The base station is switched off and can neither receive nortransmit messages

• Active Idle The base station is not operating a traffic channel, andtherefore radiates a dummy bearer that the mobile receivers are able todetect when observing the physical channels

• Active Traffic The base station operates at least one traffic channel.The dummy bearer is no longer supported

• Active Traffic and Idle The base station maintains a dummy bearer

in addition to at least one traffic channel (traffic bearer )

In the basic state Active Locked, the base station transmits a dummybearer in order to enable mobile terminals to synchronize with its frame andslot clock If a traffic bearer is established, the base station moves into Active

-Figure 9.22: Classification of MAC services

Traffic state The dummy bearer can then be eliminated The transition onthe opposite side takes place after the last traffic channel has been released If

a dummy channel is required when the traffic bearer is transmitted, the basestation can move into Active Traffic and Idle state Again the dummybearer can be retained when the first traffic channel is set up A transition isthen made from Active Idle to Active Traffic and Idle

9.4.2.3 Cell and Cluster Functions

The MAC layer functions have already been covered in the overview in tion 9.3.3 This layer is used to set up and manage the traffic channels(bearers) requested from the LLME and, when requested, to release them.The control information, which is introduced through different service accesspoints in the MAC layer, is multiplexed and added to the actual user data ineach time slot

Sec-The different services of the MAC layer are divided into two groups (seeFigure 9.22) The cluster control functions in the upper area are linked to thedata link layer through the three service access points MA, MB and MC Thecell site functions in the lower part coordinate the transition to the physicallayer The two groups offer the following individual functions:

Cluster Control Functions (CCF) These functions control a cluster of cells.Each logical cluster of cells contains only one CCF, which controls all the cellsite functions (CSF) Three different independent services are available withinthis cluster:


  P

QR N Q

N MSTUMVU

 VWVUMX QR N Q

QRR TU

QR

 VWV

 VWVUMX QR N Q

Figure 9.23: Overview of services and channels in the MAC layer

Broadcast Message Control (BMC) This function only exists once in eachCCF, and distributes the cluster broadcast information to the respectivecell functions The BMS supports a number of connectionless point-to-multipoint services, which are directed from a base station to the mobilestation The BMC works with any type of traffic channel An importantservice is the paging of mobile stations

Connectionless Message Control (CMC) All information that concerns theconnectionless service is usually controlled by one CMC in each CCF Inaddition to transmitting information from the control level of the DLClayer, the CMC also processes user data from the so-called U-plane Theservices can be operated in both directions

Multi-Bearer Control (MBC) This service comprises the management of alldata that is exchanged directly between two corresponding MAC layers

An MBC capable of organizing several traffic channels exists for eachconnection-oriented point-to-point connection

Cell Site Functions (CSF) These functions appear below the CCF services

in the MAC layer, and stand in for the respective cell Therefore severalCSFs are controlled by each CCF The following cell-oriented services aredifferentiated (see Figure 9.23):

Connectionless Bearer Control (CBC) Each connectionless bearer withinthe CSF is controlled by its own CBC

Dummy Bearer Control (DBC) Each CSF has a maximum of two dummybearers to provide a beacon function for the synchronization of the mo-bile stations in the event that no user connection exists in the cell.Traffic Bearer Control (TBC) An MBC must request a TBC for a duplexconnection.

Idle Receiver Control (IRC) This service controls the receiver in a cell when

no connection is being maintained to a user; a cell can have more thanone receiver, and consequently a related number of IRC services

9.4.2.4 Service Access Points

The cluster and cell-oriented functions include several service access points(SAP), which can be used by the MAC layer entities to communicate withthe next-highest OSI layer (data link control layer ) and the next-lowest OSIlayer (physical layer ) (see Figure 9.22) The following SAPs are availablebetween the CCF and the DLC layers:

Each cell-specific service has its own D-SAP for access to the physical layer.The MAC layer has a separate SAP, the ME-SAP, to support the lower layermanagement functions This SAP is not formally specified, and therefore has

no logical channels

MA-SAP Information from the Broadcast Message Control Service is mitted to the DLC layer over this access point Data from the BS channel aswell as control data that controls the data flow on the BS channel is transmit-ted This logical Higher-Layer Broadcast Channel supports a connectionlesssimplex broadcast service from the base station to the mobile station

trans-MB-SAP This access point links Connectionless Message Control with theDLC layer, and contains four logical channels:

The control channels CLF and CLS support a connectionless duplex servicebetween the base station and the mobile terminal A continuous service existsfrom the base station to the mobile station but not in the opposite direction.The quantity of data permitted on the CLS channel is 40 bits, which corre-sponds to the segment length of this channel The fast CLF channel has apermitted segment length of 64 bits, and the quantity of data can amount tofour times the segment length

The information channels SIN/SIP offer an unprotected/protected simplexservice from the base station to the mobile station

<>=@?ACBED <>=@?ACBGF <>=@?ACBEH <I=J?ACBCK <>=@?ACBEL <>=@?ACBGFML <>=@?ACBGFMN

Figure 9.24: DECT multiframe organization

MC-SAP The multi-bearer control unit is linked to the DLC layer over theMC-SAP Five logical channels are available for the transmission of informa-tion on data flow control and establishing, maintaining and releasing MACconnections:

The control channels CSand CF offer two independent connection-orientedduplex services The maximum throughput for the connection of a slow CS

with a segment length of 40 bits is 2 kbit/s With a data segment length

of 64 bits, the fast CF control channel achieves a throughput of 25.6 kbit/swith full slots The information of both control channels is supplied with aCRC checksum (Cyclic Redundancy Check ) that provides error detection andcorrection through retransmission with an ARQ technique

One of the two information channels IPor IN is used with each transmission

to ensure that the higher layers are offered an independent connection-orientedduplex service The IN channel is for voice transmission, with the MAC layersupplying 4 bits (X-field) of error detection protection for the information

IP channels are planned for the transmission of data with error-detection orerror-correction coding

The GF channel is a connection-oriented simplex service with a data ment length of 56 bits and is used by the U-plane of the DLC layer The MAClayer provides error detection for this channel

seg-9.4.2.5 MAC Multiplex Functions

Multiframe structure In the MAC layer the TDM frame structure mented from the physical layer (see Section 9.4.1) is superimposed logicallywith a multiframe structure A multiframe is composed of 16 individual frames(see Figure 9.24), and normally begins with the first half of the frame 0, used

imple-by the base station The end of a multiframe is used imple-by the mobile terminal inthe last half of frame 15 The number of the current multiframe is transmit-ted in at least each eighth multiframe, and is used for encryption purposes.The frame number is recognized by frame 8 because this is where the logicalchannel QT is transmitted from the base station to the mobile station (seeFigure 9.26)

z0 d387 b0

d0 s0

d64

a63 a48 a47

(*),+.-*/0 1.-*/0 2436578$ :9

;<3657:$ :9

a6 a2a3 a4

= >@? R >A  CBED*FHG<IJCK L MNPOQMO

d383 d387 b419 b423

d63

Figure 9.25: DECT time-multiplexing: D32-field

D-field Bit mapping, the compilation of the D-field for subsequent mission to the physical layer, is based on fixed rules As already described inSection 9.4.1.2, the length of a D-field depends on the type of transmissionrequested If a P32 transmission is requested, the length of the D-field is 388bits A different D-structure is used for P00, P08j and P80 operations.The D-frame of a P32 packet comprises two parts, as shown in Figure 9.25.The A-field is 64 bits long and is used for the continuous transmission ofcontrol information The B-field is available for actual user data, and in fullslot operation is 324 bits in size

trans-The individual parts of the A and B-fields are explained below:

A-field The control field contains three sections (see Figure 9.25) The A-fieldinformation of 40 bits is joined to the header of the A-field, which has

a length of 8 bits An R-CRC field with 16 bits for protection of thecontrol data forms the connection

TA (a0 a2) The 3-bit long TA field at the beginning of the headerindicates the type of A-field information (a8 a47) There arefive different logical channels, one of which always transmits data

in the A-field Here a distinction is made between the internalMAC channels NT, QT, MT and PT and the CS channel (for thecontrol data of the higher layers)

Q1 (a3) In connection-oriented transmission the Q1 bit in conjunctionwith duplex bearers carries out the quality control of a channeland serves as the handover criterion In services which containerror correction it can be used for flow control

BA (a4 a6) This part describes the qualities of the B-field In tion to normal protected and unprotected information transmission(U-type, IP or IN), expanded signalling (E-type, CF or CL) is alsopossible in certain cases

N T

C T

M T N T

C T

M T N T

C T

M T N T

C T

M T N T

C T

M T N T

C T

M T

C T

M T N T

C T

M T N T

C T

M T N T

C T

M T N T

C T

M T N T

C T

M T

N

N T

C T

M T

FP

PP

Figure 9.26: Priority of control information in a multiframe cycle

Q2 (a7) Like the Q1 bit, the Q2 bit is designated for the quality trol of a connection A combination of Q1 and Q2 constitutes ahandover criterion

con-A-field info (a8 a47) Internal MAC messages can be transmittedwithin the 40-bit long tail field This involves sending out dif-ferent control information in consecutive time slots using an E-typemultiplexer

R-CRC (a48 a63) The MAC layer protects the logical channels ofthe A-field with a cyclic redundancy check (CRC) This involvescalculating and transmitting 16 redundancy bits, which with up to

• 5 independent errors

• bursty errors up to a length of 16

• error patterns with an odd number of errors

can be detected in the A-field [26] An error-correction mechanism

is not provided in the MAC layer

The internal control data of the MAC layer is contained in the A-fieldinformation One of the following options can be selected during multi-plexing (see Figure 9.26):

NT Identities Information Using its Primary Access Rights Identifier(PARI) and its Radio Fixed Part Number (RPN), the MAC layer

of the base station generates its own Radio Fixed Part Identity(RFPI), which is transmitted to the terminal

QT System Information and Multiframe Marker This channel, usedonce in each multiframe, is only transmitted by the base station,and therefore is used for indirect synchronization on the multiframecycle Information on the technical structure of the base station aswell as on the current connection is provided

PT Paging Information This channel is the only one that a mobile tion can even receive when it is in the Idle Locked state (see

!


"$#&%(')+*, -.#/%0')+*,

-.#/%0')+*,

"$#&%(')+*,

-1#/%(')+*, "$#&%(')+*,

Figure 9.27: D-fields for the different physical packets

Figure 9.20), and contains the broadcast service (BS) of a basestation In addition to the identification number of the base sta-tion, the terminal receives important MAC layer information Forexample, blind slot data that evaluates the quality of the currentbearer is transmitted and suggestions for other qualitatively goodchannels are passed on In response to a Handover Request, thehandover areas approved by the base station are transmitted back

MT MAC Control Information Administrative tasks such as tion setup, maintenance and termination and handover requestsare carried out on this channel A Wait command is sent if there

connec-is a delay in setup A reciprocal exchange of the channel lconnec-ist canaccelerate the selection of the best channel for transmission

CT Control Information Higher Layers Either CLor CSLinformationfrom the higher layers is transmitted on this channel In otherwords, this is not an internal MAC channel

Multiplexing of control channels during transmission is carried out cording to a specific plan Different control information for the A-field,which is transmitted on the basis of a priority list, is provided in eachmultiframe (see Figure 9.26) The base station always sends its systeminformation (QT) in the eighth frame, whereas for the mobile stationeach uneven frame is reserved for the NT channel If a control chan-nel with the highest priority attached to it remains unused in a frame,control information with a lower priority can be sent

ac-B-field With physical packet P32 the 324-bit long ac-B-field, for which a tected and an unprotected format exist, follows the control field (A-field)

pro-in the D-field (see also Figure 9.27) In an unprotected transmission ofdata, as in the transmission of voice, 320 effective usage bits from theB-field are required for a data rate of 32 kbit/s The remaining 4 bits(X-field) are used for error protection by the means of CRC In a pro-tected B-field format the 4 bits are kept for error protection, whereasthe 320 effective data bits are divided into four blocks The usage bitswithin each block are reduced to 64 so that an R-CRC check sequence

28572 ')(

28592 ')(

Figure 9.28: Elements of the A and B-fields

can be formed with the remaining 16 bits The net data rate reduces to25.6 kbit/s (see Figure 9.28)

9.4.2.6 MAC Services

The different MAC services produced from the combination of different cal packets and the option for protected or unprotected transmission are listedbelow These can be carried out as symmetrical or as asymmetrical services.With asymmetrical services the rigid separation between uplink and downlink

physi-is eliminated Thus the service from a base station to a mobile station can use

a slot in the first part of a frame as well as the corresponding slot in the secondpart of the frame for the downlink It is different with services that offer aguaranteed throughput or guaranteed bit-error ratio, which then results in avariable throughput In this case the MAC layer uses an ARQ protocol Theservices of the MAC layers are listed individually in Tables 9.2 and 9.3

9.4.2.7 Types of Bearers

MAC bearers are elements that are created by the cell site functions (CSF)and always correspond to a service entity in the physical layer The following

is a list of the different bearer types:

Simplex Bearer This type of bearer is used to produce a physical channel inone direction A differentiation is made between long and short simplexbearers Whereas the short bearer only contains an A-field, the long onealso transmits a B-field next to the A-field For example, dummy bearercontrol (DBC) (see Section 9.4.2.3) reserves a simplex bearer sent bythe base station for the transmission of broadcast information

Duplex Bearer A pair of simplex bearers transmitting in the opposite tion on the physical channels is called a duplex bearer The two bearersuse the same frequency and are time-deferred by half a frame (slot pair)

direc-A traffic bearer controller (TBC) uses a duplex bearer, e.g., for a trafficchannel between the base station and the mobile station

Table 9.2: Partial list of symmetrical MAC servicesa

a S T = Service type: xd = type x double slot, xf = type x full slot, xh = type x half slot

Table 9.3: Partial list of asymmetrical MAC services

in one slot pair on the same frequency This type of bearer only occurs

in multibearer connections, and is also used for asymmetrical sion

transmis-Double Duplex Bearer A double duplex bearer consists of a pair of duplexbearers that is part of a shared MAC connection Each of these duplexbearers is produced by a TBC The two TBCs are controlled by anMBC

9.4.2.8 Types of Connections

Connection-oriented entities exist in addition to the connectionless servicesthat include broadcast services Each multibearer control unit in the MAClayer is responsible for the maintenance of a connection It controls one ormore traffic bearer control entities, which are used in the administration ofthe bearers A differentiation is made between advanced connections and basicconnections:

Basic Connections A basic connection does not have a common connectionnumber that is known to the base station and to the mobile station.Consequently only one basic connection can exist between a base stationand a mobile terminal It consists of a single duplex bearer It is possiblefor two basic connections that serve the same DLC connection to existtogether for a short time when a switch is being made in physical channel(handover)

Advanced Connections Because, unlike basic connections, advanced tions use a common number, several connections can exist between twostations The individual bearers are assigned a logical number to allow

connec-a differenticonnec-ation within connec-a connection in the MAC lconnec-ayer

Physical Connections Physical connections are not supported by MAC vices (see Tables 9.2 and 9.3) They are used for non-standardized datatransmission (see Section 9.4.3.8)

ser-When a connection is being made, all the bearers requested from the DLCmust be set up within 3 s (T200: Connection Setup Timer ) or else the con-nection is considered aborted Additional bearers can be set up or released

if necessary during an ongoing connection in order to increase or reduce theoverall transmission capacity

Advanced connections can be either symmetrical or asymmetrical Withsymmetrical connections the same services and number of bearers are used inboth directions

9.4.2.9 Connection Setup

Connection-oriented procedures within the MAC layer use two point-to-pointconnections: connection and bearer Only the pure connection is apparent tothe data link layer The bearers that each connection uses for a transmissionare managed within the MAC layer and are not apparent to the higher layers.Connection setup is usually initiated by a mobile terminal (portable-initiated )

If there is a call from the fixed network, the mobile station must first be paged

so that it is aware of it and can initiate the setup procedure

Connection Setup The sequence of connection setup can be explained with

a description of the MAC layer primitives (see Figure 9.29) The higher layers

!"

' )( 

' *(  +" 

Figure 9.29: Connection setup of a basic connection

(DLC layer) initiate connection setup (MAC-Connect Request) in the MAClayer In addition to a MAC connection endpoint identifier (MCEI) that applies

as the reference address for all subsequent primitives, this primitive contains

a parameter that specifies the requested service If the MAC layer is not able

to provide the requested services, it sends a Disconnect request to the DLClayer

The instance of an MBC entity (multibearer control ) produced by the quest receives permission from the lower-layer management entity (LLME)

re-to set up a connection between the base station and the mobile station pending on the service requested, either a basic or an advanced connection isset up

De-Bearer Setup If an MBC was established through connection setup in theMAC layer of the terminal, an attempt is made to set up the bearers requiredfor the service First the mobile station must be in the state Idle Locked,

in other words, it must be synchronized with at least one base station inthe cluster that will allow it to set up its connections The MBC producesnew instances of traffic bearer control (TBC) entities, which take over therequired bearer setups A bearer setup is based on the following process (seeFigure 9.29):

1 The mobile station sends a Bearer Request for a selected channel to aknown base station It uses the first transmission code within the 40bits of the A-field

2 The base station receives the request without any errors, and establishes

a new TBC in its MAC layer

3 The TBC in the base station requests the address of a supporting stance of an MBC entity from the lower-layer management entity

in-4 If the base station is not yet ready to transmit a confirmation of thesetup to the mobile station, it sends a Wait command The mobileterminal receives this command and likewise responds with Wait

Figure 9.30: Control plane of the DLC layer

5 After completion of the protocol process on the fixed side, the basestation sends a confirmation (Bearer Conf)

6 The mobile station receives the confirmation and directly sends Other

in the next frame

7 The base station receives the Other message, and likewise responds mediately with this primitive

im-8 The mobile station receives the Other message, and the TBC informs itsMBC that the setup has been successful (Bearer Established ind)

If an error occurs at any time during the connection setup, the setup tempt is discontinued and the setup procedure is then repeated A maximum

at-of 10 setup attempts (N200) not exceeding the connection setup timer (T200

= 3 s) is permitted

The DLC layer [10] is divided into a U-plane and a C-plane (see Figure 9.16).The C-plane constitutes the control layer for all internal DECT protocol pro-cesses relating to signalling (see Figure 9.30)

It provides two independent services:

• Data link services (LAPC+Lc) • Broadcast services (Lb)The U-plane controls the transmission of user data and offers the follow-ing services, which are accessible over its own service access points (see Fig-ure 9.31):

• LU1 TRansparent UnProtected service (TRUP)

j gZ]
kWl TU[m

TU[n

]biB 

oqprrs

LNPORQ tW g au

T5UWv Y5 

Figure 9.31: User plane of the DLC layer

• LU2 Frame RELay service (FREL)

• LU3 Frame SWItching service (FSWI)

• LU4 Forward Error Correction service (FEC)

• LU5 Basic RATe adaption service (BRAT)

• LU6 Secondary RATe adaption service (SRAT)

• LU7 64 kbit/s data bearer service

• LU8-15 reserved for future services

• LU16 ESCape (ESC) for services other than standardized servicesThe specifications for services LU3 and LU4 have not yet been finalized.Subordinate frame structures that allow direct mapping of the MAC servicesare available for each of the LUx services (see Tables 9.2 and 9.3)

Each LUx service can occur in different classes of transmission:

Class 0 No LUx transmission repeat and no sequential control in the DLClayer This means that error messages from the MAC layer are forwarded

to the higher layers

Class 1 No LUx transmission repeat; the DLC layer distributes the DLCframes in the correct sequence.

Class 2 Variable throughput with LUx transmission repeat

Class 3 Fixed throughput with LUx transmission repeat

The LUx services are described below

9.4.3.1 Transparent Unprotected Service LU1

The LU1 service can only occur as a Class 0 service, and is the most basic ofthe services Although envisaged for voice connections, it can also be used fordata services No error protection is provided FU1 is the subordinate frameservice

9.4.3.2 Frame Relay Service LU2

The LU2 service is a protected service for the transmission of data blocks(frame relay) that is accessed through the LU2-SAP It operates in a genericfield of user data, which is transferred in and out of the DLC U-plane asservice data units (SDU)

LU2 supports the reliable transport of the SDUs and also respects the SDUboundaries Three basic procedures are supported:

1 Addition of a checksum to each service data unit (SDU)

2 Distribution of the resultant data-SDU+checksum to data fields of tocol data units (PDU)

pro-3 Peer-to-peer transmission of these PDUs

A distinction is made between external and internal blocks (frames) (seeFigure 9.32):

• SDUs are external blocks

• PDUs are internal blocks

Checksum procedures The 16-bit long checksum supports error detectionfor an entire SDU frame (see Figure 9.33)

A defective SDU is not retransmitted The user can designate externalerror-handling protocols

9.4.3.3 Frame Switching Service LU3

This service should be mentioned here, but the standardization for it has notbeen completed

Figure 9.33: Field format of checksum

9.4.3.4 Service with Forward Error Correction LU4

The specifications for this service have also not yet been completed It willinclude forward error correction and possibly also an ARQ entity

9.4.3.5 Data Rate Adaptation Service LU5

The service LU5 is being introduced to support synchronous data streamswith fixed data rates (64, 32, 16 and 8 kbit/s) A differentiation is madebetween protected and unprotected transmission The protected service offersconsiderably more reliable transmission

The difference lies mainly in the use of the different MAC services: IP vices are used for the protected service and IN services for the unprotectedone This produces differences in the combination of MAC services for thedata adaptation service, which justifies the separate treatment of protectedand unprotected services In general, the principle of these services can bedescribed as follows The service supports up to three independent data chan-nels There is only a limited combination of different possible data rates Rateadaptation is carried out for the individual data connections These data con-nections are then multiplexed onto a channel An option is offered here forinterleaving the data The channel is segmented into individual frames thatare provided with control information and transmitted over the radio chan-nel The entire process is performed in reverse on the receiver side so that up

ser-to three independent data channels are again available at the service accesspoint

Table 9.4: Synchronous and asynchronous rate adaptation in the LU6 service

Output data rate [kbit/s]

9.4.3.6 Supplementary Data Rate Adaptation Service LU6

The LU6 service can only operate in combination with LU5 It offers dataadaptation for terminals that comply with the V-series This service convertsdata rates in accordance with CCITT Recommendation V.110 The rateadaptations executed in the LU5 service are presented in Table 9.4

9.4.3.7 64 kbit/s Data Service LU7

This service was specifically developed to support the 64 kbit/s data service

of ISDN Since ISDN fixed networks have a lower bit-error ratio than DECTradio connections (with ISDN a BER≤ 10−6 is expected; with DECT a han-dover is initiated at a BER = 10−3), the aim of the protocol is the reduction

of bit-error ratio This is accomplished through a combination of forwarderror correction (FEC) with an RS code and ARQ procedures An increaseddata rate and buffer storage are necessary in order to maintain the ISDN datarate of 64 kbit/s Realistically it is possible to transmit with a net data rate

of 64 kbit/s or 72 kbit/s; the buffer storage produces a delay of 80 ms (seeFigure 9.34)

The transmit buffer contains the last eight transmitted blocks The receivebuffer stores the arriving (maximum eight) blocks, thereby producing the timedelay of 80 ms If a defective block is received, it is rerequested and for eight

Figure 9.35: Content of a double slot B-field from the standpoint of BS coding

transmit/receive cycles there is a possibility to receive the block again Asystem of counters ensures that the block is written at the correct place inthe receive buffer A temporary change to the increased transmission rate of

72 kbit/s compensates for the time lost through the repeated transmission.Details of the algorithms as they are specified in the standard are discussedbelow

The data fields The B-field of a double slot (see Figure 9.35) uses a (100,94)Reed–Solomon code with 94 bytes for the ARQ mechanism and 6 bytes forthe checksum

The ARQ mechanism divides the 94 message symbols into three groups(control, information and checksum, CS); see Figure 9.36

The checksum field contains a checksum which the ARQ mechanism used todetermine whether forward error correction was successful The informationfield contains 90 bytes of user data when there is a 72 kbit/s transmissionformat; with a 64-kbit/s transmission format the last 10 bytes are filled withzeros (see Figures 9.37 and 9.38)

The control field contains the counters and format control parameters portant for the ARQ mechanism (see Figure 9.39)

im-The send and receive buffer storage and status variables V (O), V (R), V (S),

V (A) and Vi(T ) can be seen in Figure 9.40 These regulate the ARQ process

by indicating which frame has been sent, received or acknowledged and from





Figure 9.37: ARQ information field for 64 kbit/s user data rate

Figure 9.38: ARQ information field for 72 kbit/s user data rate

Figure 9.39: Content of ARQ control field

which storage position in the send storage read should take place or at whichstorage position in the receive storage write should take place The param-eters N (O), N (R), N (S) transmitted in the ARQ control field and formatsderived from the status variables are used in the communication between ARQentities

ARQ Control Fields

Format field The transmission format is coded with a total of 4 bits thatare divided over two 2-bit fields (format 1, format 2); see Figure 9.39.The receiver must be informed of which transmission format is beingused so that it can interpret the last 10 bytes as user data or zeros (seeFigure 9.37)

ReTransmit requests (RTR) are sometimes also sent in these fields ble 9.5 shows the significance of the bits

Ta-The Offset variable V(O) indicates how many additional bytes (in units of

10 byte) are still required in order to refill completely the receive buffer

of the partner instance It also determines the respective transmissionformat (64 kbit/s or 72 kbit/s)

When a frame is retransmitted, V (O) is increased by 8 (corresponding

to the deficit of 80 bytes in the occupation of the receive buffer to becompensated for) Retransmission is only permitted as long as V (O)≤

48 If V (O) is greater then retransmission is not possible because thedata requested is no longer available in its entirety in the transmit bufferand, furthermore, would no longer have a chance of arriving in time at

of the base station generates its own Radio Fixed Part Identity(RFPI), which is transmitted... beginning of the headerindicates the type of A-field information (a8 a47) There arefive different logical channels, one of which always transmits data

in the A-field Here a distinction is made... data bits are divided into four blocks The usage bitswithin each block are reduced to 64 so that an R-CRC check sequence