Tse Ss2004 Notes

8/6/2019 Tse Ss2004 Notes

1/322

I

Stephan RuppGerd Siegmund

Telecommunication Software Engineering

Lecture NotesEdition: V 0.2, December 20, 2006

Copyright 2004 by Stephan Rupp and Gerd SiegmundReproduction is not permitted without the consent of the authors.

Contact: [email protected]: http://www.srupp.de


2/322

II


3/322

IIIForeword

Foreword

This manuscript originates from lecture notes of joint and individual activities of the authors in 2004.Also, it is based on material the authors have used in other publications. The topic has been inspired

by the opinion, that telecommunication is becoming a core discipline of information technology,along with data bases or operating systems.

Telecommunication software represents a natural follower of the microprocessor: Most systemsdo not operate in isolation, but are in communication with others. What information technology can

benefit from telecommunications are solid engineering methods and the design of complex systems.Engineering practice in telecommunications today incorporates new technologies such as Web-Engi-neering, novel applications of data bases and storage systems, as well as the object oriented frame-works and methods for the development of applications. While this goes well beyond the traditionaldesign of functional architectures, it is still rarely reflected in textbooks about engineering methodsin telecommunications.

This manuskript intends to address software engineering in a modern telecommunication envi-

ronment. It is organised in 6 chapters, which are largely self contained and may be read indepen-dently from each other. In the same way, single chapters may be used as material to supporting trai-ning or lectures.

Chapter 1 is a summary of telecommunication networks and mobile networks. Someone familiarwith this may directly proceed to chapter 3, which is about distributed computing. Chapter 6 is aboutsecurity. Chapters 1, 3 and 6 represent an entry level of the material. The other parts move deeperinto specific subjects: chapter 2 covers protocols and protocol engineering, chapter 4 is about ser-vices architectures within the network and on terminal devices, and chapter 5 explains methods forsoftware engineering. The exercises are organisaed in the same way: Part 1 of the exercises followthe entry level tour, part 2 covers the complete manuskript.

In total, the focus is on basic concepts and on methods. In chapters 1 to 6, concepts and methodsare developed close to reality, or at least shown in a practical or realistic context. The exercises fol-low a more conventional approach. They stay on an abstract level, but also are aimed to show the ba-sic concepts and methods.

Acknowledgements: Leonhard Stiegler has made available the practical case for protocol con-formance tests in section 2.4. The summary of the Bluetooth protocol stack in section 2.5 has been

provided by Harald Orlamuender, who also has contributed on CORBA and Web-Services in section3.2.5. Chapter 6 is based on a large extent on material provided by Matthias Duspiva. Xuepeng Tanhas provided the material on the development environment in section 5.3. The service architecture


4/322

ForewordIV

which is presented in chapter 4 contains many creative ideas from Rodolfo Lopez-Aladros andFranz-Josef Banet.Thanks to all of them for their contributions and for frequent discussions. KlausJobmann of the institute for Communication Technologies (IANT) at the university of Hannover hasencouraged us on the service architecture, we thank for his continuous support. Many thanks to PaulKhn and the Institute for Communication Networks and Computer Engineerings (IKR) at the uni-versity of Stuttgart for hosting the lecture and for their contributions on modelling and simulation.

Stuttgart, July 2004

Stephan Rupp Gerd Siegmund

Note on Edition 0.2, December 2006: After the 3rd term, the exercises have been updated. Theexercises include the examinations within the first 3 years of the lecture. Each examination consistsof two tasks with maximum 8 questions according to the following logic:

Exercises 7.2.14 and 7.2.15: Trial examination 2004Exercises 7.2.16 and 7.2.17: Examination 2004Exercises 7.2.18 and 7.2.19: Examination July 2005Exercises 7.2.20 and 7.2.21: Examination October 2005Exercises 7.2.22 and 7.2.23: Examination July 2006Exercises 7.2.24 and 7.2.25: Examination October 2006

Stuttgart, December 2006

Stephan Rupp


5/322

VHow this book is organised

How this book is organised

Chapter 1 - Telecommunication Networks

The first chapter represents a summary on commu-

nication networks with emphasis on 2G (GSM, GPRS)and 3G (UMTS). It explains the general structure of tele-communication networks and what happens, when a mo-

bile phone is switched on or when a call is initiated. Alsoit explains the design principles of telecommunicationnetworks and of services in telecommunication net-works.

Chapter 2- Protocol Engineering

When terminals or network elements communicatewith each other, they are exchanging messages in a spe-cific context and according to specified rules. Chapter 2

explains the concept of such protocols at different levelsof abstraction.The chapter contains the ISO/OSI Refe-rence model, state diagrams, data flow diagrams, the ba-sics and a practical test for protocol conformance testingand a summary of the the Bluetooth protocol stack as a practical example.

Chapter 3- Network Organisation

How may functions be separated and allocated todifferent network elements or terminal equipment?Chapter 3 explains the difference of distributingfunctions within one system and distributed systems,summarises the corresponding technologies, and ex-

plains the basic principles of service discovery and andorganisation of networks. It also introduces the perspec-tive of an application programmer on connections and on using middleware to handle distribution offunctions in order to develop services.

?

?

?

?


6/322

How this book is organisedVI

Chapter 4- Service Architectures

What do modern terminals such as mobile phones andnetwork elements provide for the development of ser-vices and applications? Chapter 4 has a look at traditio-nal functional design, design while using a service archi-tecture based on distributed computing, as well as theapplication frameworks available at modern terminal de-vices. At abstract level, chapter 4 explains how resources

for computing, storage and network capacity are handled. It then shows a representative practicalcase for a service architecture (a virtual HLR/VLR) and the application framework of a mobile

phone (based on Symbian operating system). Chapter 4 introduces class diagramms and message se-

quence diagrams.

Chapter 5- Engineering Methods for Service

sWhile Chapter 4 is focussed on what the network ele-ments provide for software development, Chapter 5 has alook at the methods for software development and at thedevelopment process. The basics of how to apply theUniversal Modelling Language (UML) are shown bysupplementing the diagrams which have already beenshown in Chapter 2 and 4 by introducing use cases, acti-

vity diagrams, and package diagrams. Concerning the development process, the components neededfor a prototype development (development environment and runtime environment) are shown. WhileChapter 5 cannot provide a comprehensive view of tools and process, the target is to give an ideaabout the practicalities of software development.

Chapter 6- Security for Services and Applications

While we increasingly carry communication deviceswith us and are inceasingly living in an environmentready to communicate with us, security with respect tothe design of software and the handling of software is

becoming increasingly important. Chapter 6 explains thesecurity threads and countermeasures within the networkas well as for terminals. It introduces the basic

prionciples such as active attacks, passive attacks, protective measures, sandboxes for software, si-gned content, secure connections and how cryptography is applied.

Exercises Part 1 - Entry Leve

lThis book may be used in two levels: Chapters 1, 3 and 6

provide the entry level (i.e. without going deeper intoprotocols and software engineering, but covering the ba-sic principles about networks, distribution of functionsand security). Part 1 provides the corresponding exerci-ses.

??

software

?software

Chapter 1 Chapter 6

Chapter 3


7/322

VIIHow this book is organised

Exercises Part 2 - Complete Manuscript

Part 2 of the exercises provide a deeper level inclu-ding protocols, systems design and the basics of softwareengineering. For self-study, it may be recommendable todo Part 2 after Part 1 of the exercises. For training or lec-ture, the sequence does not really matter.

Web-Site of the book

It is planned to provide the illustrations, exercisesand more supporting material on a web-site. For the time

being, the material used in the lecture of 2004 can be re-

trieved from:

http://www.ind.uni-stuttgart.de/Content/Courses/Over-view/SS-2004.html

Chapter 1 Chapter 6

Chapter 3Chapter 2 Chapter 5

Chapter 4

WWW


8/322

How this book is organisedVIII


9/322

- V -

Table of Contents

1 Telecommunication Networks 1

1.1 Introduction 1

1.1.1 Circuit switched networks and packet switched networks 2

1.1.2 Access networks and core networks 5

1.1.3 Services and applications 9

1.2 Terminology 12

1.2.1 Communication model 13

1.2.2 Connections 14

1.2.3 The architecture of telecommunication networks 151.3 Concepts for Switching and Routing 18

1.3.1 Connection orientet exchange of messages 18

1.3.2 Connectionless exchange of messages 19

1.3.3 Circuit switched connections 19

1.3.4 Store and forward routing 22

1.3.5 Message switching (datagram service) 23

1.3.6 Cell based connections (ATM) 23

1.4 Mobile communication networks 24

1.4.1 Summary of mobile communication systems 24

1.4.2 Basics of radio transmission and cellular networks 26

1.5 GSM - Global System for Mobile Communication 28

1.5.1 Basic architecture 28

1.5.2 Services 291.5.3 Establishment of connections 30

1.5.4 Interfaces and Protocols 33

1.5.5 Addressing and Identification of subscribers 34

1.5.6 Sequences in calls and connections 36

1.6 GPRS - General Packet Radio Service 41

1.6.1 Summary 41

1.6.2 Architecture 42

1.7 UMTS - Universal Mobile Telecommunication System 43

1.7.1 Architecture 44

1.7.2 UMTS Phase 1 (Release 3 or Release 99) 46

1.7.3 UMTS Phase 2 (Release 4/5) 47

1.7.4 UMTS Service Areas 481.7.5 Identities and Security Aspects in UMTS 49

1.7.6 UMTS Services 50

1.8 Web based service architectures 51

1.8.1 Traditional architectures for services 52

1.8.2 Object oriented software design 54

1.8.3 Web-Services and distributed computing 56

1.8.4 A challenge for a new service architecture 60

2 Protocol Engineering 63

2.1 Processes and Protocols 63

2.1.1 ISO/OSI Reference Model 64

2.1.2 Numbers and Addresses 70

2.1.3 Communication Processes and Protocols 71

2.2 Formal Specifications 73


10/322

- VI -

2.2.1 Formal Methods 73

2.2.2 Finite State Machines 75

2.2.3 Specification and Description Language (SDL) 75

2.3 Formal Verification 77

2.4 Implementation and Tests 79

2.4.1 Hardware and Software Design 80

2.4.2 Model of a protocol stack 812.4.3 Protocol Tests 82

2.5 A practical case: INAP conformance tests 84

2.5.1 Conformance testing concepts 84

2.5.2 INAP interface test configuration 85

2.5.3 Protocol Tracer Configuration 92

2.5.4 Sample Message Trace 93

2.6 The Bluetooth Protocol Stack 95

2.6.1 Bluetooth Radio Technology 95

2.6.2 Networking 97

2.6.3 Protocols 99

2.6.4 Profiles and Service Discovery Protocol (SDP) 104

3 Network Organisation 107

3.1 Basic Principles 107

3.1.1 Communication and Context 107

3.1.2 Service Discovery and organisation of networks 113

3.2 Some Concepts in Practice 116

3.2.1 JXTA networks 117

3.2.2 Bluetooth Service Discovery 119

3.2.3 Java RMI and JINI 121

3.2.4 J2ME MIDP 2.0 125

3.2.5 CORBA and Web-Services 127

3.2.6 Universal Plug and Play 131

3.3 Export and Import of Functions 131

4 Service Architectures 133

4.1 Introduction 133

4.1.1 Resources: Storage, Computing, Network 134

4.1.2 Centralised Data in Communication Networks 135

4.1.3 Framework Protocols 140

4.2 Network Service Architecture 144

4.2.1 Concept 145

4.2.2 Data Interface 147

4.2.3 Application Interface 151

4.2.4 Consolidation of Interfaces 152

4.3 Data Models and Semantic Models 154

4.4 Terminals and Embedded Systems 1544.4.1 Symbian Architecture 154

4.4.2 Client and Server Threads 157

4.4.3 GUI Framework 160

4.4.4 Data Storage 162

4.4.5 Communication Framework 163

4.4.6 Messaging Framework 169

5 Engineering Methods for Services 173

5.1 Introduction 173

5.2 Engineering Services with UML 174

5.2.1 A sample application 174

5.2.2 Summary of popular UML diagrams 180

5.2.3 Tests 1815.3 Development Tools and Process 182


11/322

- VII -

5.3.1 Development Environment 182

5.3.2 Runtime Environment 184

5.3.3 Development Cycles 185

5.4 Plug-ins and Bundles 187

6 Security for Services and Applications 189

6.1 Sandbox and Middleware 1896.1.1 The Java Sandbox as role model 189

6.1.2 Middleware - the J2ME MIDP security concept 191

6.1.3 Buffer overflows 193

6.2 Common IP security issues 195

6.2.1 Wireless network access 195

6.2.2 Security Issues with IP networks 195

6.2.3 Security issues with wireless personal devices 197

6.2.4 Technical solutions 198

6.2.5 Procedures to address security issues 204

6.3 Identity 206

6.3.1 Identity, Authentication and Authorisation 206

6.3.2 Implications on Accounting 209

6.3.3 Security and Privacy Requirements 210

6.3.4 Credentials 212

6.4 Web Services Security 214

7 Exercises 215

7.1 Part One (Chapters 1, 3, and 6) 215

7.1.1 System Availability 215

7.1.2 Mobile Terminating Call 215

7.1.3 Virtual HLR/VLR 217

7.1.4 Data Traffic in Wide Area Networks (WAN) 218

7.1.5 Mobility 219

7.1.6 Location based Services 221

7.1.7 Distributed Computing 2227.1.8 Service Inventory 223

7.1.9 Name, Address and Identity 225

7.1.10Name Spaces 225

7.1.11CORBA and RMI 227

7.1.12Clients and Servers 228

7.1.13Threads and Processes 228

7.1.14Memory Leak 229

7.1.15Viruses and Trojan Horses 229

7.1.16Service Discovery 229

7.1.17Bluetooth 229

7.1.18Communication End Points 229

7.1.19File-Sharing 2297.1.20CORBA and Web-Services 230

7.1.21Direct Provision of Client Applications 230

7.1.22A Sandbox for Software 230

7.1.23Buffer Overflow 230

7.1.24The phantom menace 231

7.1.25Attacks 231

7.1.26Requirements on public network infrastructure 231

7.1.27The security hole in your pocket 231

7.1.28Town Gate 231

7.1.29Defence System 232

7.1.30Key Ring 232

7.1.31Session Keys 232

7.1.32GSM Authentication 2327.1.33One-Time Pads and Stream Ciphers 232


12/322

- VIII -

7.1.34IPSec 233

7.1.35IP-VPNs 233

7.1.36Identity, Authentication, Authorisation 233

7.1.37Biometrics 234

7.1.38GSM-Keys 234

7.1.39Cloned SIM Cards 234

7.1.40Credentials 2347.2 Part Two (Complete Book) 234

7.2.1 Collisions in Piconets 234

7.2.2 Framing protocol 236

7.2.3 Protocol Overhead in VoIP 238

7.2.4 Shopping carts 239

7.2.5 Output Buffer 239

7.2.6 Intelligent Network 240

7.2.7 Virtual HLR in a Mobile Network 240

7.2.8 Location Updates 241

7.2.9 Classes and Objects 241

7.2.10Relations at Examination Time 241

7.2.11Evaluation of Results 242

7.2.12Service Discovery 245

7.2.13Personal Networks 248

7.2.14Mobile Instant Messaging 250

7.2.15UPNP Remote Control 254

7.2.16Location Based Services 256

7.2.17Client-Server Communication 259

7.2.18Local Context Service 261

7.2.19Home Networks 264

7.2.20Fixed Mobile Convergence 266

7.2.21Mobile Peer-to-Peer Networks 270

7.2.22Mobility Management 272

7.2.23Remote Control for Home Networks 277

7.2.24Mobile Video-Streaming 2797.2.25Remote Player 281

8 References 285

9 Translations English - German 289

10 Abbreviations 291


13/322

1.1 Introduction 1

1 Telecommunication Networks

In this chapter, we will have a look at telecommunication networks and at theservices they provide. Section 1.1 represents a qualitative summary of net-works. Section 1.2 will present a look at the terminology used in communica-tion networks. Section 1.3 provides the essential concepts of circuit switchedand packet switched exchange of information. Section 1.4 summarises the es-sentials of mobile communication networks. The following sections 1.5 to 1.8introduce the basic architecture of existing network technologies (that is GSM,

GPRS, UMTS and the World Wide Web). In perspective, chapter 1 is intendedas introduction to communication systems as basis for telecommunicationsoftware engineering.

1.1 Introduction

world wide connec-

tions ...

The world wide telephone network represents the biggest machine whichhas ever been build by humankind. However, little of it is known to us in eve-ryday life. Who knows what is behind the telephone outlet, the cable modemor DSL modem? What makes mobile phones work?

Connections to the world wide communication infrastructure is becomingan elementary feature in our households, in education, in business, and incre-

asingly while we are on the move. Most probably, information technology willchange the way we work much faster and fundamentally than the steam engineor the car have been doing over the last centuries. Following work, capital andsoil, information is becoming an increasingly important asset in our society.Communication networks carry information and this represent a vital infra-structure in our society.

... change the way we

live

Globalisation of markets and customisation of products will influence theway products are designed and manufactured. Most probably, it will also fun-damentally change the way we work. Working time and the place of work arebecoming flexible. This enables people to get organised in a flexible way, insmaller teams with flat hierarchies and short communication links.

In many cases, communication networks already are a basic condition formany jobs (how many enterprises do operate without telephone or PC?). Tele-communication facilitates co operation with other people, wherever they are.


14/322

2 1 Telecommunication Networks

It also facilitates trends such as the current migration of jobs to start-up coun-tries.

Fig. 1-1

Telecommunication

networks carry digital

information (bits)

What are telecommunication networks about? Figure 1-1 shows a summary.Communication networks cannot carry physical goods. They carry data, suchas digitised media. Media includes any kind of information, such as words, si-gnals, voices, music, pictures, movies or files and messages exchanged bet-ween machines. The digitised information is represented by numbers, or bits.For transport, bits may be packaged into packets if information. The most re-levant (the most universally accepted) packet format today are IP packets.

1.1.1 Circuit switched networks and packet swit-ched networks

The traditional way of connecting to ends to each other has been a circuit. Inanalogue networks at the beginning of last century, the circuit also representedan electrical circuit for the signal. Essentially, the idea of a circuit is to connectto ends through pipes. Figure 1-2 shows the principle.

Fig. 1-2

Switched connections

- signalling phase


15/322

1.1 Introduction 3

Between the connected parties, the network has generated an individual pi-pe. The pipe is used for a stream of information (usually in both directions).Usually, the flow of information is possible with a guaranteed continious datarate (such as 64kBits per second by a telephone line or ISDN B-channel).

Fig. 1-3

Circuit switched con-

nections - connection

phase

Todays circuit switched networks are completely digital, such as the tele-phone network that handles analoge terminals and ISDN, or the GSM net-works for mobile phones. The circuits does not exist in a verbal sense, but asa concept:

- the network establishes a communication channel- the channel provides a constant bitrate- the channel is maintained until one of connected parties hangs up.

The concept of packet switched network is equally simple, as shown on Figure1-4. It corresponds to packets deliverd by a postal service. Packets are labelledwith the adress of their destination (and the address of their sender). The postalservice takes care of routing and delivery.

Fig. 1-4

Packet networks


16/322


The implications on the delivery of the specific content of the packets areless simple. Packets may carry an e-mail, an audio stream or video stream, ora telephone call. Requirements on latency for each service are entirely diffe-rent: a video stream may arrive 10 seconds later without bothering the user.The natural interactions in a telephone conversation need latencies below 50milliseconds (you will need to interact with "roger and over", if latency goesbeyond 200 milliseconds). For a download of a file or for an e-mail, latency isin the order of minutes or hours. In order to carry such different loads, packageswitching networks will need a concept to provide different qualities of ser-vice.

Fig. 1-5

Example: Voice over

Packets

How does the transfer of information through packages work? Figure 1-5shows the principle. As example, the information is represented by a voiceconveration with the user on the left talking to someone on the right. At thesender, the signal needs to be digitised, coded and cut into piexec of a givenlenght. These pieces are put into packets, together with a time stamp. This pro-cedure inevitably generates a delay (or latency).

Packets are now transferred over the network. Depending on how manyhops the packed needs to pass, and depending on traffic conditions on the net-work, this will generate another delay.

Jitter translates into

delay

At the receiver, the pieces of information are lined up according to their ti-mestamp. This procedure is called "dejittering" (the jitter represents variationsin the delay caused by the transfer over the network). Also, the signal needs tobe decoded before it is transformed into an audible signal at the speaker of thephone. The procedures at the receiver also cause further delays. In particular,the delay caused by dejittering corresponds to the delay caused in the network(this is well known from streaming audio or video over the internet, where themedia player buffers the signal in order to generate a continuous output sig-nal).


17/322

1.1 Introduction 5

Fig. 1-6

How to deliver packets

over a large distance?

A brief look at packet transport: Does it make sense to deliver each packetindividually? This will depend on distance and on the volume of traffic. For aPizza-Service, it may be o.k. to deliver orders individually. For a postal ser-vice, distance and volume make other solutions more efficient. Figure 1-6

shows the principle.

Postal service

Packets from one area to another area are collected in a container and trans-ferred together. At the destination, the container is uncharged and packets arerouted to their destination. At physical level, wireline infrastructure or wire-less infrastructure may be used.

1.1.2 Access networks and core networks

Following the qualitative view of some basic concepts in the last section, wewill now have a look at the geographic proportions of networks and their con-sequences on network technologies. We will start with larger distances.

A road system for data

traffic

To connect the larger cities in a place like Germany, we think about the mo-tor highway or the intercity railway network. In terms of telecommunication

networks, this corresponds to the data highway, which is represented by opti-cal fibre networks. Figure 1-7 shows such a network.


18/322


Fig. 1-7

The information high-

way

However, this information highway does not carry regional traffic or metro-politain traffic. In terms of the railway system, this corresponds to everythingthat a regional train or metro line (S-Bahn) connects. In terms of the road sys-tem, this also corresponds to a regional or metropolitain network. In a metro-politain communication network, optical fibre connects the main sites, such asdata centres, main distribution frames of the telephone networks, fibre nodes(hubs) of Cable TV networks, or base station controllers of mobile networks.

The distribution networks which connect to the metropolitain fibre connectto the homes and offices of the end users, the so called last mile. Figure 1-8shows the metropolitain area and two squares, which represent cuts of about 5kilo meters and 500 meters. Depending on the density of population, a wirelineor wireless access network would supply such an area.

Fig. 1-8

Metro networks and

the last mile


19/322

1.1 Introduction 7

As example for networks on the last mile, Figure 1-9 shows a cable TV net-work. Cable TV networks originally represent long trees of coaxial cable andcoaxial amplifiers. In order to allow interactive services (such as Internet ac-cess via cable modem or telephone services over cable), areas within the cop-per network are supplied via optical fibre. Figure 1-9 shows the correspondingfibre nodes (hubs) and the coaxial trees that emerge from them. Dependingupon demands of bandwidth, one area represents about 5000 housholds whoshare the cable. Also shown in Figure 1-9 is a photograph of a stree cabinetwhich holds a fibre node of caxial distribution. Similar streat cabinets are usedto distribute copper wires of the telephone networks (or electrical power).

Fig. 1-9

Example: Cable TV

networks

Figure 1-10 shows a summary of the different types of networks. All net-

works shown represent wide area networks (WAN). They include access net-works in the last mile, fibre rings in metropolitain areas, and the data highwaysto bridge national and international distances. At access level, the major phy-sical infrastructure is represented by copper wires (of the telephone network),CaTV networks, and the base stations for cellular mobile networks. By theproportion of connections they provide, access networks represent the biggestshare of investment in network infrastructure (digging cable and mounting ba-se stations on a large scale is expensive).

Fig. 1-10

Fig. 1-10 Summary of

network infrastructure


20/322


Network design fol-

lows traffic

Given the high density of traffic in a densely populated area, fibre rings con-nect the different access networks and also collect their traffic. Transmissionsystems used in fibre networks allow to carry both packed switched and con-tinuous types of traffic. The fibre networks in the metropolitain area connectto the access points of different network providers, which are shown as Pointsof Presence (POP) in Figure 1-10.

The point of presence is providing network access (it generates the User-ID/Password request of your internet connection or allows your mobile phoneto use the network). It also represents a point of interconnect between differentnetworks and network operators. In other cases, such as an IP based cable net-work connecting to a circuit switches telephone network, it performs a conver-sion of media (such as packets to circuits). The POP shown in Figure 10 is anabstraction of different network elements which are used in different network

technologies to perform such functions.At backbone level, network elements which switch telephone traffic in GSM

networks or PSTN networks, or handle large amounts of packets in the inter-net. The performance of today's network elements allows to handle highlyconcentrated traffic. For instance, a mobile switching centre can handle milli-ons of mobile subscribers. A switch (exchange) in a public telephone exchangecan handle hundreds of thousands of subscribers. Before the digitisation ofnetworks, there had been 8000 analogue telephone exchanges in Germany (fora total of about 40 millions of telephone lines). 20 years ago, they have beenreplaced by 1500 digital exchanges. With today's technology, about 100 sys-tems would be sufficient.

Fig. 1-11

Fig. 1-11 What about

Satellites?

So far, satellites have not been mentioned. Figure 1-11 shows a scanrio,which includes satellites in the communication networks. Because geostatio-nary satellites cover a large area, they are particularly usesful to broadcast in-formation to a large number of destinations. For the same reason, they are lessuseful to generate a individual communication links between a large numberof users.

Digital transponders allow to broadcast any type of information (includingthe Internet) to different destinations. The bandwidth covered by one TV chan-nel (analogue channels with 5 MHz each) corresponds to about 35 Mbits of da-


21/322

1.1 Introduction 9

ta (or 6-8 digital TV channels). Media received by satellite may also may bedistributed over wireline access networks, such as telephone lines (over DSL).Not shown in Figure 1-10 are terrestrial TV networks. Terrestrial TV networksare now also being digitised and may supplement wireless networks (such asUMTS).

Fig. 1-12

What about Wireless

LAN and Bluetooth?

Local Area Networks

With the exception of satellites (and terrestrial TV), which directly reach theend user, the other networks discussed so far basically represent Wide AreaNetworks (WAN). In the local area, there are private telephone systems with private networks, or PCs and servers connected in a Local Area Network(LAN). Still closer to the end-user are the devices he may connect to with theBluetooth interface or infrared interface of his PDA, mobile phone or PC. Suchnetworks are called Personal Area Networks.

Figure 1-12 shows a summary. It should be noted that local area networks(wireline or cordless telephone systems or wireline or wireless data networks)usually connect to the subscriber interface of a Wide Area Network (such as aDSL modem, cable modem or equivalent interface).

1.1.3 Services and applications

Services generate traf-fic

Services are functions that networks provide to us (such as a mobile phonecomes with a telephone service). Applications are something that can be pro-vided on top of the basic services (such as an interactive game that can bedownloaded and installed on a mobile phone). Increasingly, networks becomeopen for new applications. Both services and applications are the essential partof Telecommunication Software Engineering. In this section, we will raise so-me questions on services and applications.


22/322


Fig. 1-13

Where do we use net-

works?

First of all: Where do we use networks? As shown in Figure 1-13, networksare largely used at the desk of an office and home office. The mobile phonehas enabled us to use networks everywhere. Further services will be providedon the move (by adding data services to the mobile device which allows us toarrange our life or work on the spot), as well as in the living room (basicallyentertainment, but also home automation).

Fig. 1-14

What services do we

use?

What services do we use? Figure 1-14 provides a classification with someexamples: conversational services, interactive services, streaming services andservices which may run in the background. Again, services are moving fromthe traditional combination of telephone and PC into different areas. They aregetting more diversified. Also, networks will go beyond connecting people:


23/322

1.1 Introduction 11

they will connect an increasing number of machines and vehicles. For net-works, it will become important to ba able to handle a diversity of services.

Fig. 1-15

The value chain: Who

is providing which ser-

vice?

Who is providing services? As shown in Figure 1-15, there is also a growingdiversity of players in the value chain. In the traditional telephone network, thenetwork operator provides the complete service (sometimes also including theterminal set).

For new services, the value chain is more complex. It largely corresponds tothe current practice in the internet or in the entertainment media business: A

content provider provides the media (such as a web-site or movie), a portalprovider aggregates different offerings of content (such as Yahoo or iTunes),a service provider specialises on end user demands (such as providing Internetaccess, web-hosting, IT-services, hotlines). In this scenario, the role of the net-work operator is reduced to transport of data (and the traditional telephone ser-vices).

Fig. 1-16

Who will use future

services and which

concepts are used to

provide services?


24/322


Services are used by

people and machines

Who will use future services? Which concepts will be used to support suchservices in communication networks? Figure 1-16 shows a summary. Users to-day are largely people. Tomorrow, networks will connect an increasing num-ber of vehicles, machines, appliances, consumer electronics, sensors and ac-tors.

The predominant concept to implement services in networks today is the cli-ent-server architecture. Concepts tomorrow include software downloads & up-dates, user agents, peer-to-peer networking, and ad-hoc organisation of net-works and resources. Applications will become more data centric, diverse anddynamic.

This development is largely driven by the migration of the microprocessorinto almost every device, i.e. the increase of so called embedded systems.Maybe, in a couple of years, we will see a microprocessor in every lightswitch.

A natural follower to the microprocessor is a communication infrastructure.This is because one thing that the microprocessor will certainly demand is asoftware update. Generally, it will also want to communicate the state of thesystem it represents or take instructions from a controlling network element.

The conclusion is, that we will have two major areas of development in te-lecommunications: (1) networks for embedded systems, (2) service architec-tures in core networks.

Fig. 1-17

How are networks or-

ganised?

After covering some qualitative views, there are two remaining questions:how are networks organised and is there is an abstract and systematic way tohandle communication networks? This will be shown in the following sec-tions.

1.2 Terminology

Networks, Access

Points, and Terminals

The main task of a communication network is to allow the exchange of mes-sages between the access points of the network, i.e. to transport messages fromone access point to another access point or to other access points. Figure 1-18shows the principle.


25/322

1.2 Terminology 13

Fig. 1-18

Network Access

Points

The main requirements on such a network are:- an access point must be available at any time to transfer messages to

any other access point- the user at an access point is able to decide on the destination of the

message- the network must be able to handle a large number of simultaneous

transactions- the technical efforts to do this is minimised by utilising the statistical

qualities of the transactions.

Telecommunication networks operate in large scales: The telephone networksfor instance connect more than 1 billion telephone subscribers and more than1.5 billion mobile phone subscribers worldwide.

1.2.1 Communication model

Terminals and Appli-

cations

Users do not directly communicate with network access points

. They areusing terminal devices like a telephone or a desktop computer. The terminaldevices provide the user interface

(towards the user), as well as the user net-work interface (towards the network). Figure 1-19 shows this basic communi-cation model.

Modern terminal devices like an ISDN phone or a desktop computer do notconnect directly to the network access points, but are using intermediate net-

works like the S0 bus or a LAN-connection to another device, which connectsto the network access points (such as a ISDN NT or a DSL modem or cablemodem). In terms of Figure 1-19, all of this represents terminal equipment (so-metimes also called customer premise equipment because it is installed at thecustomer's place).

Source/Sink

Source/Sink

Network

Network

Access Point


26/322


Fig. 1-19

Communication Model

Messages

are created by the applications running in the terminal devices.Among such applications are the telephone service, FAX, e-mail and all appli-cations running on a desktop computer which communicate over the network.The terminal equipment at the user network interface (such as modems) trans-forms messages the respective types ofsignals

to transfer them over a largerdistance to the network access points. Such transformations happen repeatedlywithin the network.

Services to support

applications

In order to enable applications, the network provides services. For instance,

the telephone service implemented in the telephone networks enables voicecommunication as application. The telefax service enables the exchange offacsimile documents as application. The e-mail service enables the exchangeof electronic documents. The World Wide Web enables access to documentsstored in the network.

1.2.2 Connections

Post-Box or on the line

A connection

represents the establishment of a communication channel bet-ween the communicating parties for a period of time. The communicationchannel may be used to exchange messages. There are different types of com-munication channels, which the network can provides. For instance, an ISDNconnection offers a channel for the transparent exchange of messages at a con-tinuos bit rate of 64 kbits per second. A completely different type of connec-tion is represented by e-mail. In this case, the communication partners connectto post boxes. The network transfers messages between the post boxes. Figure1-20 shows the path of a connection through the network.

Source/

Sink

Source/

Sink

Network Access

Network

Application

Message Transfer

Signal Transfer

User

Interface Network Interface

Termination Termination

Network Interface

User

Interface

Points


27/322

1.2 Terminology 15

Fig. 1-20

The route of a commu-

nication link through

the network

Allocation of resources

for connections

There are many ways of implementing connections. The traditional way hasbeen the physical connection between communication partners by a switch-board. The modern equivalent are connections at continuous bit rates providedby time division multiplex systems (such as ISDN mentioned above). Anotherconcept are virtual channels, which provide the required bitrate on demand andelse allocate resources to other traffic flows. Connections may be provided onrequest of the user (dial-in connections), or be provided between known com-munication partners continuously (fixed lines). In general terms, a connection

exists, as long as the communicating partners maintain a communication link,i.e. they are in a state of communication with each other.

1.2.3 The architecture of telecommunication net-works

Hierarchical structure

How do networks manage to provide connections between a large numberof access points? Figure shows the principle architecture of a network such asthe telephone network. It is using an hierarchical structure. There are networkelements (nodes), which connect to network access points (with terminalequipment attached to them by access lines). In a telephone network, such net-work elements are called

local exchanges

.

calling

Party called

Party

Connection

Verbindungs-

leitung

Destination

Address

Communicaton Channel

Network Node,Exchange

Telecommunication

Access-line

AB

Network


28/322


Fig. 1-21

Structure of a telecom-

munication network

efficient utilisation of

resources

These network elements at the lower network level connect to network ele-ments further up the hierarchy. In a telephone network such network elementsare called transit exchanges

. Network elements at this level do not carry sub-scriber lines, but connect to each other. Also shown in Figure 1-21 is a connec-

tion through the network. The network takes care that the route of such a con-nection is short and utilises resources in the best way. In a telephone network,a path is chosen before the connection is established (this happens only afterthe called person picks up the phone). If, for instance, the called person is busyor does not accept the call, no connection is established at all (in this case, youwill hear a busy signal, or hear a free signal).

Network Layers In reality, networks are a bit more complicated than shown in Figure 1-21.They are using a layered approach which allows an optimum utilisation of re-sources. These layers are:

- Transmission Layer- Switching and Routing Layer- Service Layer.

Figure 1-22 show an overview. The transmission layer provides connection to

access networks. It also represents connectivity to switching networks (bothfixed networks and mobile networks) and the Internet. In switching network,a service layer has been separated in order to facilitate network design, plan-ning and operation. The service layer provides services such as number trans-lation and special tariffs by the so called Intelligent Networks (such as free-phone 0800 or premium rate services such as 0900). It also provides numberportability.

Connecting Lines

Access Line

Communication Link

Switched

Terminal Sets

Bundle (Channel)

Bundle

Connection


29/322

1.2 Terminology 17

Fig. 1-22

Traditional Network

Structure

InternetThe Internet is by nature an overlay network which may utilise any existingnetwork infrastructure (it is an inter-network network). Internet network ele-ments such as Routers and the Domain Servers may be connected anywhere inthe network infrastructure. However, the volume of traffic provided by Inter-net users and the number of users using services like the WWW and e-mail hasmade the Internet a driving force in network design and planning.

Broadband access such as DSL-Modems or Cable Modems are specificallydriven by the Internet. Also, the provision of services in the WWW has gene-rated extremely effective engineering methods (Web-Engineering), that willdrive the provisioning of services and applications. Similarly, the wide accep-tance and universal connectivity of IP as routing protocol is driving a next ge-neration of network elements. In such networks, IP is becoming a universaltransport layer. Figure 1-23 shows the principle architecture.

Intelligent Network

SCP

IN

Transmission Network

Internet

Access-Netz

Access Network

Mobile-Network

Fixed Network

Intelligent Network

SCP

IN

Scitching Plane

Service Plane

Transmission Plane


30/322


Fig. 1-23

Separation of Trans-

port and Control in

Next Generation Net-

works

Besides the prominent role of IP as transport layer (over existing transmis-sion technologies), the most significant difference between Figure 1-23 andFigure 1-22 is the separation of control of traffic and handling of traffic. Thetraffic is handled by gateways (which for instance pick up a 64 kbit channel,transform it to packets which it sends to another destination).

Next Generation Net-

works

The traffic is controlled by gateway controllers (which for instance instructa gateway to switch a connection to a specific destinations). The IP-layer ofthe network (routing layer) in such a network has to handle a variety of trafficclasses (such as e-mail, media streaming and interactive services).

1.3 Concepts for Switching and Routing

1.3.1 Connection orientet exchange of messages

The communication in connection oriented networks follows a general se-quence:

- connect: The network uses the destination address provided by the user,

Intelligent Network

SCP

Call Feature-ServerSignalling & Media Gateway Controller

IN

Control

Internet + Transmission Network

Access Network

Access Network

Mobile NetworkFixed Network

MultimediaApplications

Intelligent Network

SCP

IN

MultimediaApplications

Call Feature-ServerSignalling & Media Gateway Controller


31/322

1.3 Concepts for Switching and Routing 19

finds out whether the destination is ready to communicate (i.e. notbusy), reserves the required resources and establishes the connection.

- use the connection: The network maintains a communication channelfor the exchange of messages between communication partners.

- disconnect: Upon request of one of the communication partners, thenetwork de-allocates all resources and cancels the connection.

Quality of ServiceOne important part of this sequence is the allocation of resources. It allowsto provide a specific quality of service, once a connection is established. Re-sources will be available to the communication partners as long as theymaintain the connection. For a regular telephone call, the quality of service re-presents a continuous bit rate, a specific quality of the voice transmission (suchas echoes, delays, distortion). Such quality criteria are strictly specified and

need to be maintained by suppliers of infrastructure and by operators, who in-terconnect their networks.Also, the network is able to manage it's resources and avoids degradation of

services. For instance, in new years night, telephone networks tend to get un-der heavy traffic conditions. They operate at maximum load and accept onlythe number of calls they can actually handle (so if you receive a busy signal onyour mobile in such a situation, it is not because the network has crashed, butbecause it is fully loaded).

1.3.2 Connectionless exchange of messages

Transfer of packetsComputers connected to a LAN or over the Internet do not follow a se-quence of establishing a connection (at least not at Ethernet level or IP level).

In a connectionless environment, a device is allowed to send a packet contai-ning the message spontaneously. The packet contains the destination addressand the address of the sending device (so that the destination may respond).

Packets are transferred through the network according to their addresses.This transfer is also called routing. In order to be able to do this, routers willneed to maintain information about the network topology. However, they donot allocate resources. In an overload conditions, packets are allowed to getlost. The benefit of such networks are that they only need resources when the-re are actually packets to transfer (they combine different packet streams in astatistical way, which is calledstatistical multiplexing).

Differentiation of trafficHowever, this situation makes it difficult to maintain a quality of service insuch networks. Delay times are unpredictable and may have significant varia-tions. Heavy load conditions are difficult to handle. This situation correspondsto cars and vehicles on a highway. On a highway, the faster vehicles are allo-wed to use a dedicated lane. Differentiation and priorisation are also conceptsto improve quality of services in connectionless networks.

1.3.3 Circuit switched connections

In this section, we will have a closer look on how connections work in a circuitswitched network. This essentially explains, what happens when you pick upyour phone and dial a number.


32/322


Establishing a connection

Figure 1-24 show the basic sequence of events. As soon as subscriber A picksup the phone, he indicates the wish to establish a connection. If the network isready to accept a connection, it quits by sending a dial tone. The subscribernow dials a number (or selects a number from the telephone directory on thephone). The network is checking, whether the destination B is ready to accepta call (e.g. not busy, or mobile phone not switched off), then allocates resour-ces and lets the phone at B ring.

Fig. 1-24

Sequence of events in

a telephone call

The networks sends a free tone to subscriber A, which indicates that the pho-ne on the other end is ringing. As soon as B picks up the phone, the networkestablishes a communication link. A and B are nor connected and may ex-change information. In the scenario shown in Figure 1-24, the connection ismaintained until A hangs up, which is indicated to B. Terminating the connec-tion includes de-allocation of all required resources by the network.

Signalling

Separation of control

and payload

What Figure 1-24 does not shows is the communication between the net-work elements which is associated with the connection. Telephone exchangesuse special protocols and channels for such procedures, which are called sig-

nalling. Signalling is used between terminal devices and network accesspoints, as well as between nodes in the networks, in order to exchange controlinformation. In classical networks designs, control is entirely separated by theuser information and even uses separate channels (such as the ISDN D-chan-nel) or even networks (such as the signalling system 7). Separation of control(Control Plane) and user information (User Plane) is also maintained in nextgeneration network architectures and ist associated protocols (such as SIP).

Network elements

How does a telephone exchange actually look like? Figure 1-25 show the prin-

A Off-Hook

Dial Tone (Accept)

Dial Digits

Ring Tone

Ring Signal

B Off-HookB Off-Hook Indication

Exchange of User Information

Release (B On-Hook) Release Indication (Busy)

Release (B On-Hook)

User A User BNetwork

Connection

Establishment

Connection

ReleaseTime


33/322


ciple architecture of a local exchange. On the left, it connects subscriber lines(via line cards). On the right, it connects to other exchanges (over so calledbundles, which are each going to the same destinations). At the core of the ex-change, there is the digital switching matrix, which allows to connect user traf-fic from different incoming ports to different outgoing ports, as well as thecontrol.

Fig. 1-25

Structure of a local te-

lephone exchange

Control information is handled separate from user traffic. This is indicatedin Figure 1-25 by signalling channels leading from subscriber line terminati-ons to the control box, which is using the information provided through signal-ling to transfer signalling information to other exchanges and to control istown switching fabric (and vice versa for terminating calls).

Circuit switched con-

nection

Figure 1-26 again summarises the complete scenario of a connection fromA to B. In a circuit switched network, the characteristic part is the connectionpoint maintained in each switch (telephone exchange), which is part of theconnection. What circuit switched network also maintain is information aboutthe distance and duration of connection. Such information is used for accoun-ting and billing.

Bundle Direction1

Direction 1

Direction N

Bundle DierectionN

Trunks for

InterconnectionLine

Terminations

Switching Network

Control Processor

Control

Local Telephone Exchange

.

.

.

.

.

.

.

.

.

Signalling Signalling


34/322


Fig. 1-26

Circuit Switched Con-

nections

1.3.4 Store and forward routing

The transfer of messages over a packet based network may follow an entirelydifferent principle. Because packets have a limited duration, they may bestored in a network element and then transmitted again in the appropriate di-rection. A condition to do this is, that each packed contains a label which in-dicates control information about the destination or logical connection the pa-cked belongs to. Figure 1-27 shows the fragmentation of a message into apackage.

Fig. 1-27

Fragmentation of a

message into packets

Each packet consists of a header (which contains addressing information),and a body (which contains the message or part of it). In messages fragmentedinto several packets, the sequence of the fragments also needs to be included(for example by a sequence number).

Packets are generated at the source of the communication network. A paththrough the network may be shared by several connections, by transferring thecorresponding packets in an alternating way (one using the pauses of theothers). Through storing and forwarding packets, the network elements repre-sent a buffer (or queue) for packets.

A

Calling

Party

Connection

Communication

Channels

(Bundles)Network SwitchingNode (Exchange)

Telecommunication Network

AccessLine

B

Physical Connectio(Crosspoint)

Crosspoint

Link

Message Blocks:Message

Header

Body

Packet


35/322


Fig. 1-28Packet Switched Con-

nections

Figure 1-28 shows the principle. For the user, a packet switched connectionlooks like a permanent connection. Still, resources are only allocated if theyare actually used to transfer a packet, i.e. during the period of time a packed isstored within each network element. In practice, packets with the same desti-nation will always take the same path (i.e. as long as the routing informationin the network elements is stable and does not change). By nature, packet swit-ching does not exclude the notion of a connection in terms of a common pathand destination of a sequence of packets.

1.3.5 Message switching (datagram service)

In message switching, each packet represents a complete and independentunit, i.e. a complete message or datagram. Each network element will store andforward the packet entirely based on the information included in the header.There is nor relation between packets, even if they originate from the samesource. Else, message switching entirely corresponds to the store and forwardrouting described in 1.3.4. A practical example for message switching are IPpackets or UDP packets in the Internet.

Fig. 1-29 Generation

of a packet at the

sender

1.3.6 Cell based connections (ATM)

A conveyor belt for

data

One property of packet switched networks makes mixing of different kindsand qualities of traffic difficult to mix with each other: variable and unpredic-table packet lenghts. This makes it difficult for continuous types of traffic(such as a sequence of small packets each carrying 10 mseconds of a voicecall) to mix up discontinuous traffic with long packet lengths (such as from adownloaded file).

A

Network

BufferNetwork Switching Node

B

Switching

and forwardof packets

by storage

Message Block:Body Header

Header: Control (destinationl, lenght, priority...)

Body: Payload


36/322


Fig. 1-30

Fragmentation of a

message into cells

Cell based connections like ATM are using a stream of packet entities witha fixed length, the so called cells, as shown in Figure 1-30. As usual, a messageis fragmented into the cell stream. The cell stream represents a continuous

flow with predictable resources trough the networks elements (see Figure 1-31), even if cells are empty. At the incoming ports of each network element,cells are processed according to the information in their header and put on theappropriate outgoing cell streams.

Fig. 1-31

Cell based connec-

tions

Because it allows to mix up different kind of traffic in combination with anefficient use of resources, ATM is used as transport layer to aggregate IP traf-fic and TDM traffic in access networks, as well as for exchanging and distri-buting traffic in core networks.

1.4 Mobile communication networks

Like mushrooms Mobile communication networks have had a remarkable start-up world-wi-de. In Germany, for instance, it has taken less than 10 years for mobile phones

(from the availability of the first digital network in 1991) to grow beyond 50million mobile subscribers, which is higher than the number of fixed telephonelines (it took about 100 years to grow a number of about 40 million fixed tele-phone lines). Also, mass production has enabled mobile phones to becomesmall and ubiquous. Beyond cellular mobile networks, other wireless techno-logies are emerging.

1.4.1 Summary of mobile communication systems

The most simple mobile communication is using a cordless phone. As indica-

Message

Cell Header

Cells

A

Network

Header Processing

B


37/322


ted by the name, the cord between the handset and the telephone set is replacedby a radio interface. All handsets associate with the fixed line that their basestation connects to. They either associate with this number or, in the case of aprivate exchange, use an extension of this number. Cordless phones are sup-ported by the same network as fixed line and do not cause any changes there.

Where ever you areIn comparison, mobile cellular networks, which handle mobile phones (pre-dominantly networks using the GSM standard), are quite complex. They allowsubscribers to roam (that is use their phone in any place world wide). They alsomanage to keep a conversation going while the user is on a train or in a car (i.e.they need to hand over the radio connection between different base stations wi-thout interruption). In order to identify subscribers and to encnrypt communi-cation, they use smart cards (SIM cards).

In the last couple of years, GSM has been extended by a packed based data

service, GPRS (General Packed Radio Service). GSM has been designedabout 20 years ago as the second generation of mobile networks which has re-placed former analogue networks. Over the coming years, the next generationof mobile networks will be deployed, the third generation: UMTS (UniversalMobile telecommunication System). In fact UMTS claims to be more than thenext cellular system. It also promises integration of other technologies to pro-vide a uniform level of service.

Fig. 1-32 Mobile

Communication

Technologies

Figure 1-32 shows a summary of the different wireless technologies. It indi-cates the bandwith they provide versus the degree of mobility. For instance, acordless telephone technology may provide indoor mobility, but fail to providestationary mobility (roaming or nomadic mobility) or movement between base

UMTS

Cordless

(CT, DECT, WPABX, WLL) Cellular Systems

WLAN

0,01

0,1

1,0

10

100

Transmission Rate

[Mbit/s]

Office/Home Building stationary walk drive

Indoor Outdoor

wireline terminals

(GSM and others)

Mobility


38/322


stations (handover). The same holds for wireless LAN (unless it is integratedas radio access technology in a UMTS network). In this section (and the follo-wing sections until 1.8), we will focus on the cellular mobile networks and theservices they provide.

1.4.2 Basics of radio transmission and cellular net-works

Tough conditions for

radio waves

Radio waves do not enjoy the comfort to travel in a predictably environmentsuch as waves travelling along through a fibre or along a pair of copper wires.Between sender and receiver, much more difficult conditions need to behandled. Radio waves suffer attenuation and reflections. In an urban area, re-

ception without multiple reflections would be quite difficult, as shown in Fi-gure 1-33. In a rural area, there are far less reflections (except for mountains,which cause rather long delays of about 20 microseconds compared to 0,2 mi-croseconds in urban conditions). Multiple reception (but also absorbtion, po-larisation and interference) is causing fading of signals received.

Fig. 1-33

Radio Reception

Cellular networks Cellular networks also are designed to handle fast moving users (i.e. theDoppler shift of frequencies while moving). GSM is designed to support mo-ving users up to speeds of 250 km/h (which corresponds to a fast moving caror train). But why are those networks called cellular?

The idea is to provide a large area of coverage by a grid of cells which clo-

sely fit to each other. Figure 1-34 shows the principle: each base station covers3 areas (cells), which intersect with other cells to a cluster. In theory, the cellsrepresent hexagons. The cluster shown has a similar structure than a single celland may be used this way to generate larger structures. The different shades inthe figure represent different frequencies. Frequencies may be reused in a re-asonable distance, which results in a tile structure. Interference between cellsis further minimised by power management, i.e. the base station limits the po-wer of the mobile set to the required strength.


39/322


Fig. 1-34

A Cluster of nine Cells

The size of a cell depends on the power of the base station. The size is ad-justed to the traffic density, which corresponds to the density of the population.In an urban area, we should expect smaller cells than in a rural area. In GSM,cell sizes may vary between a diameter of 250 m and 35 km (with 6 to 8 km

on average).

Fig. 1-35

Size of Cells corres-

ponds to Traffic Densi-

ty

Figure 1-35 shows cells covering an urban area, a suburban area, as well asa rural area including a highway.

A

A

A

B

B

B

C

C

C

Location of theBase Station

Rural

Motorway

Suburban

City

Area

Area


40/322


1.5 GSM - Global System for Mobile Commu-nication

1.5.1 Basic architecture

No national bounda-

ries

From the beginning, GSM has been planned as a seamless digital Europeanmobile network system, that does not end at the national boundaries. The spe-cifications have been published by ETSI in 1990. In the subsequent years,GSM has been deployed in Europe and now provides a seamless connectivityfrom Spain to Sweden. Also, it has gained world wide acceptance and now re-presents the predominant cellular network technology.

Figure 1-36 shows the network architecture. Geographical coverage is ob-tained by base stations which generate radio cells. Adjacent cells do not inter-fere with each other, so a seamless coverage can be achieved. Two basic sub-systems of GSM can be differentiated:

theNetwork Subsystem (NSS), which represents an ISDN like network ar-chitecture, and is shown on the right of Figure 1-36 beginning with theMobile Switching Centres,

theRadio Subsystem (RSS), which represents the base station controllers,bas stations and the radio cells on the left of Figure 1-36.

Fig. 1-36

Structure of a GSM

Network

The Network Subsystem connects the mobile network to public telephonenetworks (via Gateways Mobile Switching Centres). Mobile Switching Cen-

Visitor

Visitor

Radio Cell

BTS

Register

Register

Home Register

Authentication

Equipment Reg.

Authority

Home Register

Authentication

Equipment Reg.

Authority

Signalling Channelswith Mobile Applicaton

Protocols

Signalling Channelswith Mobile Applications

Part and ISDNApplications

Protocols

Tepehone

MSC

GatewayMSC

BSC

BSS

Network


41/322


tres (MSCs) are the equivalent of telephone exchanges. However, they do notcontain line cards to connect subscriber lines. Else, the mobile network usesthe same circuit switched principles as in fixed networks (see section 1.3.3).

Mobile users may

show up anywhere

However, there is an entirely new aspect with mobile subscribers: a mobilesubscriber may show up anywhere in the network (respectively, in any mobilenetwork). So the network needs to identify the subscriber, accept outgoingcalls from the subscriber and deliver calls to the subscriber.

This is achieved by adding additional network elements: the Home LocationRegister (HLR) is a data base which holds the basic profile of the subscriberand associates with the subscribers mobile telephone number. The Visitor Lo-cation Register (VLR) contains information about subscribers which are actu-ally roaming in the area covered by an MSC, i.e. dynamic data. The role of theHLR is that of the local post office in the home town of the subscriber. While

the subscriber has temporarily moved to another place, it will know the postoffice of the visited place (the VLR) and arrange for traffic to be forwarded.Other network elements allow to identify SIM-cards and equipment (mobilesets). How the different components inter-operate, will be shown with somesequences in section 1.5.3 and 1.5.6.

In total, the Network Subsystems consists of the following components:- Mobile Switching Centre (MSC)- Visiting Location Register (VLR, usually part of an MSC)- Home Location Centre (HLR)- Authentication Center (AuC, usually part of the HLR)- Equipment Identification Register (EIR, not mandatory)- Operation and Maintenance Centre.

The Radio Subsystem connects the mobile sets over the radio network to themobile switches. It include the mobile terminals (i.e. mobile stations and hand-held devices), the base stations and the base station controllers.

The last two components form another entity in GSM: theBase Station Sub-system (BSS) . The BSS includes:

Base Station Controllers (BSC)

Base Transceiver System (BTS, i.e. the base stations).

Base Stations (BTS) and Base Station Controllers may be collocated (i.e. at thesame location) or operate remotely (i.e. separated from each other). A BSCmay connect to several base stations.

1.5.2 Services

Mobile telephone ser-

vice

The basic service that GSM provides is telephony. The telephone service al-so includes the display of calling numbers and other features (so called supple-mentary services) from ISDN. However, the physical implementation is diffe-rent. Over the air, GSM is using speech compression and channels with maxi-mum bit rate of 9,6 kbits (which makes the quality of GSM inferior to what weknow from a fixed line).

In order to minimize the use of radio resources, a voice activity detector eli-minates pauses in conversation from transmission. This discontinuous trans-


42/322


mission also helps the terminal set to save power. GSM coded signals aretranscoded into 64 kbits bitstreams at connections to public telephony net-works (and vice versa).

FAX and data The use of voice compression make FAX services and analogue modem ty-pes of services difficult to directly connect over GSM. However, digital dataservices are supported with 9,6 kbits per channel. The mobile set may be usedas modem to connect a notebook computer to the Internet (via serial interfacecable, infrared or bluetooth).High Speed Circuit Switched Data (HSCD) al-lows to bundle several channels together in order to achieve higher bit rates.

SMS Another service which had been designed into GSM from the beginning isthe Short Message Service (SMS). Today, it has reached an astonishing accep-tance and popularity. SMS allows to send text messages of up to 160 charac-ters per message, which are transported in the signalling network of the GSM

Network Subsystem-

1.5.3 Establishment of connections

The typical assemply of components of a GSM network are shown in Figure1-37. On the left, the mobile set connects to a base station over the GSM airinterface. As indicated on the top of Figure 1-37 , the Base Station Subsystemconnects to the Network Subsystem. There, the other components mentionedabove con be found.


43/322


Fig. 1-37

GSM Components

The mobile set, which today is a rather small device of less than 100 gramsof weight, still indicates its origin by the correct name: mobile station. Literal-ly, a "mobile station" is a portable radio station. Because subscribers in a mo-bile network cannot be identified by the line or radio channel they use, the mo-

bile set needs to provide its specific identity to the network.Portable radio stationsThis is achieved by the Subscriber Identity Module, the well known SIM-

card. The SIM card is a chip card (or smart card), which contains a secret key.This key is used to identify the subscriber (actually the key is never transferredover the network but used to generate a signature to a random message send tothe mobile set, which proves that the mobile set knows the secret). In order toavoid misuse, the SIM card may be protected by a PIN code, which must beentered by the subscriber in order to activate the card (and thus the mobile set).

OMC

MSC

BSC

digital radion transmissionBTS

BTS

MS

A-Interface

Abis-Interface

Air-Interface

BSS NSS

AC: Authentication Center

BSS: Base Station Subsystem

BSC: Base Station Controller NSS: Network SubsystemOMC: Operation and Maintenance Center

MSC: Mobile Switching Center

MS: Mobile Station

EIR

VLR

AC HLR

BTS: Base Transceiver Station

VLR: Visitor Location Register

EIR: Equipment Identification Register

HLR: Home Location Register

Voice Mail System

SMS-C

SMS-C: Short Message Service Controller


44/322


Fig. 1-38

Basic Sequence of a

Mobile Terminating

Call

What happens, when

you get a call?

In order to show how the different network components co operate, we willfollow the sequence of an incoming call from a fixed line to a mobile subcri-ber. Figure 1-38 shows the different steps:

(1) The call from the fixed network is transferred to a Gateway MSC.The correct mobile network and the Gateway MSC may be identifiedfrom the structure of the mobile number (see section 1.5.5)

(2) The Gateway MSC requests information about the actual location ofthe subscriber from the HLR

(3) The HLR delivers the required information. If the subscriber is notlocated in the area of the HLR, the HLR knows which VLR actuallytakes care of the subscriber.

(4) Using the information obtained from the HLR, the MSC initiates aconnection to the target MSC.

(5) The target MSC reuests the actual location of the subscriber from itsVLR.

(6) The VLR delivers the requested coordinates.

(7) The MSC does know which area, but not exactly in which cell thesubscriber is in. So it pages all BTS in the corresponding area.

(8) The mobile set responds on the page requests, authenticates its iden-tity and receives a session key for encryption. The connection is establis-

Mobile

Telephone

Radio Cell

Radio Base

6

6

6

7

7

7

8

65

3

2a

2

14

Location Area

SCGateway-

Mobile

SC

VisitorRegister

HomeRegister

HLR

VLR

2b

Network

Station


45/322


hed as soon as the subscriber answers the call.

1.5.4 Interfaces and Protocols

The sequence described in the last section demonstrates, that, most probably,there are many protocols to be followed and a variety of messages to be ex-changed. In this section, we will have a look at interfaces and protocols. Figure1-39 shows the corresponding view of Figure 1-37.

Fig. 1-39

Interfaces and Proto-

cols

Behind the curtainThe interfaces shown in Figure 1-39 are:

Um: represents the interface between mobile station and base station, i.e. the

air interface, which uses TDMA-Frames (Time Division Multiple Ac-cess) in a GSM specific format

Abis: uses 2Mbits/s lines to connect the base station to the base station con-troller. The traffic from the radio part per 2 Mbit/s line represents up to80 channels (in a 64 kbits/s structure with 16 kbits/s per channel)

A-Interface: compliant with regular 2Mbits/s trunks in switched networksincluding the signalling system no. 7. Each traffic channel now repre-sents the standard format of 64 kbits per channel.

The protocols shown in Figure 1-39 correspond to three layers: physical layer(layer 1), data link layer (layer 2) and network layer (layer 3). First, we have a

Layer 2LAPDm

Layer 2LAPDm

Layer 2LAPD

Layer 2MTP

Layer 2MTP

Layer 1 Layer 1 Layer 1 Layer 1 Layer 1

Layer 3SCCP+ MTP

Layer 3SCCP+ MTP

BSSAPBSSAP

RR

Layer 2LAPD

Layer 1

RRRR

MM MM

CC SMS SSCC SMS SS

Um Abis A

Mobile BTS BSC MSCSet

Air Interface Interface Interface


46/322


look at the layer 2 protocols:

LAPDm: A mobile version of the ISDN framing protocol LAPD, whichtakes care of the safe transport of signalling information

MTP 2: Message Transfer Part of the signalling no. 7 type of layer 2 proto-cols.

The layer 3 protocols differentiate the different functions of the network ele-ments in GSM. Functions such as Radio Resource Management (RR) need tobe contains in the mobile set, the base station and the base station controller.Other functions such as Mobility Management (MM) are just passed by th-rough BTS and BSC (see Figure 1-39), they concern arrangements betweenthe Mobile Set and the Mobile Switching Centre (MSC). A brief summary of

the layer 3 protocols (or rather sub layers of layer 3):Radio Resource Management (RR): This protocol arranges the resources of

the air interface, such as access to channels and allocation of channels. Italso takes care of searching for a subscriber (paging) for an incoming call(see step (7) in the scenario in section 1.5.4), as well as random access tochannels by a mobile set for outgoing calls.

Mobility Management (MM): As already stated, this protocols concerns ar-rangements between the mobile set and the MSC. Mobility Managementincludes procedures for Authentication, which is the basis of the locali-sation of a roaming subscriber. It also arranges for the hand-over proce-dure of a moving subscriber to another cell. Mobility Management is thecondition for the following protocol layers (sub layers).

Call Control (CC): Establishes, monitors and disconnects incoming andoutgoing connections. Co-operates with the following SupplementaryServices and Short Message Service. Together, these protocols arenamed Connection Management.

Supplementary Service support (SS): Handles special features such as dis-play of the called party number, put calls on hold, hand-over of calls, 3party conference, display of charging information etc.

Short Message Service (SMS): Handles exchange of telegrams between mo-bile sets (internally represented as part of the signalling messages).

1.5.5 Addressing and Identification of subscribers

How to address a mo-bile user?

With moving subscribers through the network, addressing and identificationis much more complex than in a fixed network. Subscribers are addressed bytheir mobile telephone number and also own a universal identity, which isstored in their SIM card. Home Location Register (i.e. the home post office)and Visitor Location Register (i.e. the post office at the visited place) take careof the required associations between addresses and identities.

Addresses:

Mobile Subscriber Number (MSISDN): is associated with a mobile termina-tion in a specific GSM network. The format is compatible with the inter-national scheme for telephone numbers (the so called E.164 numbering


47/322


scheme). Figure 1-40 shows the structure. A Mobile Subscriber Numbercontains a country code (49 for Germany), a national destination codewhich represents the network (or is translated into a network), the codeof the HLR in charge of this number, and finally the subscriber number.Thus, the MSISDN is a pointer to the HLR which is in charge of a sub-scriber.

Mobile Station Roaming Number (MSRN): A temporary number assignedby the VLR and associated with a visiting mobile set, which allows toaddress the VLR.

Handover Number: From the perspective of the MSC in charge of a call,handover represents a call forwarding (i.e. the MSC stays in charge ofcall control). The handover number is used to forward a call to another

MSC in case the subscriber is handed over into the radio domain of thisMSC.

Fig. 1-40

Mobile Subscriber

Numbers

Who is this?Identities:

International Mobile Subscriber Number (IMSI): A universal unique iden-tifier for the mobile subscriber, which is stored on the SIM card. If themobile set is switched off (or the SIM card is removed), the MSC willblock incoming calls (or divert them to a mail box) in order to avoidpointless paging).

Temporary Mobile Subscriber Identity (TMSI): Once a subscriber is authen-ticated (i.e. his identity has been proven), the VLR in charge assigns atemporary identity to this subcriber. The temporary identity works as apseudonym for the subscriber and helps to protect his privacy.

Local Mobile Subcriber Identity (LMSI): An optional further local pseudo-nym which may be assigned by the VLR following updates of the sub-scriber location.

International Mobile Station Equipment (IMEI): Not only the SIM card hasa universally unique identity, but also (and independently from the SIMcard) the mobile set: the IMEI. This basically corresponds to the use ofMAC addresses in computer networks. The IMEI may be used to locateor to bar stolen devices.

Identities for location areas and base stations: In order to support handovers, there are identities for location areas. Other geographical identi-

4 9 171 - D1

172 - D2

177 - E+

179 - E2

HLR Subscriber

Number (SN)

Country

Code

National

Destination

Code HLR Destination Code


48/322


ties are cell identities, which identify a radio cell within a location area.Base station identity codes allow to identify base stations from differentnetworks.

1.5.6 Sequences in calls and connections

In this last section about GSM, we will show some scenarios for outgoingcalls, incoming calls, location updates and hand over in more detail. The ideais to give an impression what is specific for a mobile call, and what the asso-ciated protocols are doing. It will not provide an in depth analysis.

Outgoing call

Radio channels are not assigned to terminals, but represent a shred mediawhich needs to be allocated according to demand. For an outgoing call, themobile set will first need to get hold of a channel (by using a "random access"procedure), and then signal a service request. Figure 1-41 shows the sequenceof events.


49/322


Fig. 1-41

Sequence of an outgo-

ing mobile call

BSS VMSC/VLR GMSC Exchange

Random Access

Establish a Signalling Link

Service Request

(MS-No.,Service type ...)

Encryption

(Cipher Mode Setting)

Authentication

(RAND/SRES)

SETUP

(B-Number, Bearer Service, Tele Service...)

Allocate User Channel

IAM

(B-Number, Bearer Service,Tele Service...)

ACM

(B-User available, Ringing)

CONNANS

(establish connection)

Cennection

ALERT


50/322


Next, the mobile set will need to identify (this procedure is called authenti-cation). Following authentication, encryption is set up with a session key sendby the network. At this stage, the regular procedure for setting up a call con-nection follows (which includes the allocation of a traffic channel and alertingthe called party).

Incoming call

What, more specifical-

ly, happens, when you

get a call?

The incoming call (see sequence shown in Figure 1-42) to a mobile subscri-ber corresponds to a mail which is relayed to the actual place where the sub-scriber is located. First of all, the subscriber number (MSISDN) allows to iden-tify the correct network and the responsible HLR (home post office). The HLR

contacts the VLR/VMSC in charge, which generates a transfer of the tempora-ry address (MSRN, see section 1.5.5) to the MSC.


51/322


Fig. 1-42

Sequence of an inco-

ming mobile call

BSS

Set up

Paging

IAM

(B-Number,

ACM

(B available, Ringing)CONN

ANS

(establish connection)

Tele Service)

MAP: Send

(B-Number

HLR RequestTele Service )

Routing Info

MAP: Provide

Check authorisation andrequested services

(MSRN)

MAP: Result

(MSRN)

MAP: Result

IAM

(B-Number,

Tele Service)

(IMSI/TMSI)Random Access

Signalling Connection

Response to Paging

(MS-No./CKSN/MS-Classmark)

Authentification

Connection Resolution

Cipher Mode Setting

Call Setup

(BC/Facility)

Assign Channel

(TCM Assignment)

Connection

ALERT

HLR GMSC ExchangeVMSC/VLR

Roaming Number

(IMSI/LMSI)

Read subscriber datafrom VLR


52/322


The GMSC now sets up a connection to the visited MSC (VMSC). Still theexact location of the mobile subscriber is not known (i.e. which radio cell heis in). Thus, the target MSC (VMSC) is paging the corresponding base stati-ons. The requested mobile set responds by starting the random access proce-dure to allocate radio resources. After a signalling connection between mobileset and VMSC, is established, the mobile set responds to the paging request.

The following steps are not entirely new. Following authentication and setup of an encrypted connection, a call is set up and the calling party is alerted(that the mobile set is ringing now). If the called party accepts the call, a con-nection is established.

Location management (Roaming)

How does the network

keep track of mobile

users?

How does the Home Location Register know, which Visiting Location Re-gister is handling the subscriber? This is achieved by a procedure called Loca-tion Update. Location updates take place if the mobile set is switched on, whe-never it is moving into a different location. They also take place, when a mo-bile set is switched on in a new place (e.g. following arrival at an airport orrailway station). Figure 1-43 shows the procedure.

Fig. 1-43

Location Updates

In GSM, a location update is always activated by the mobile set. If the mo-bile set moves into the area of a new VLR, the HLR is informed about thechange. There are two different cases here:

(1) The mobile set is not registered in any VLR in this are (e.g. if it isswitched on after a travelling for a while). In this case, the subscriber in-formation is requested from the HLR and transferred to the VLR in char-ge.

(2) The mobile set has been registered in an VLR already. In this case,the subscriber information is transferred from the previous VLR to the

HLRBSS VMSC/VLR

new

Random Access

Establish Signalling Connection

Request Location UpdateUpdate Location

Insert

Subscriber Data

Subscriber Data

Insertion Ack.

Update Location

Ack.Confirm Location Update


53/322

1.6 GPRS - General Packet Radio Service 41

new VLR in charge.

A location area may compris

Tse Ss2004 Notes

Documents

Transcript of Tse Ss2004 Notes