Positioning_EDGE.doc

TIK-109.551

Research Seminar on Telecommunications Business II

POSITIONING EDGE IN THE MOBILE NETWORK

EVOLUTION

12.3.2003

Vesala Sami 47278H

Koivu Katja 44217E

HELSINKI UNIVERSITY OF TECHNOLOGY

DEPARTMENT OF COMPUTER SCIENCE

Tik-109.551 Positioning EDGE in the mobile network evolution Spring 2003

1

Table of Contents

TABLE OF CONTENTS 1

ABBREVIATIONS 1

1. INTRODUCTION 4

2. EVOLUTION FROM GSM TO GPRS NETWORK ARCHITECTURE 5

2.1 CELLULAR PLATFORM EVOLUTION 5

2.2 GSM 6

2.2.1 Functioning of the GSM system 7

2.2.2 Architecture of the GSM network 7

2.2.3 GSM mobile stations and SIM cards 13

2.3 HIGH SPEED CIRCUIT SWITCHED DATA 13

2.4 GENERAL PACKET RADIO SERVICE 14

2.4.1 GPRS in general 14

2.4.2 GPRS Architecture 16

2.4.3 Logical Channels of GPRS 17

2.4.4 GPRS coding schemes 18

2.4.5 GPRS terminals 19

2.5 EVOLUTION OF GSM DATA SERVICES TOWARDS EDGE 20

2.5.1 Short Message Service 20

2.5.2 Wireless Access Protocol 21

2.5.3 Multimedia Message Service 21

2.5.4 HSCSD and GPRS enabled services and data rates in practice 22

3. EDGE TECHNICAL FUNDAMENTALS 24

3.1 8-PSK MODULATION IN GSM/EDGE STANDARD 25

3.2 ENHANCED GENERAL PACKET RADIO SERVICE (EGPRS) 26

3.2.1 Link adaptation 26

3.2.2 Incremental redundancy 27

3.3 ENHANCED CIRCUIT SWITCHED DATA (ECSD) 27

3.4 EDGE EVOLUTION TOWARDS GERAN REL´5 28

3.4.1 GERAN Rel´5 features 29

3.4.2 GERAN Rel´5 system architecture 29


2

3.5 MODIFICATIONS TO THE GSM NETWORK IMPOSED BY EDGE 30

3.5.1 EDGE radio network planning compared with GSM/GPRS planning 31

4. VENDORS EDGE STRATEGIES 33

5. SERVICES ENABLED BY EDGE 36

6. TERMINAL AVAILABILITY 39

7. INVESTMENT COSTS AND REVENUES CAUSED BY EDGE 41

8. EDGE INVESTMENT STRATEGIES 44

8.1 GSM OPERATORS WITHOUT 3G LICENSES 44

8.2 GSM OPERATORS WITH 3G LICENSE 45

9. FUTURE ROLE OF EDGE 47

10. CONCLUSIONS 50

References 52


1

Abbreviations

3GPP 3rd Generation Partnership Project

8-PSK Octagonal Phase Shift Keying

AMR Adaptive Multi Rate

ARPU Average Revenue Per User

ARQ Automatic Retransmission Request

ATM Asynchronous Transfer Mode

AuC Authentication Center

BER Bit Error Rate

BLER Block Error Rate

BSC Base Station Controller

BSS Base Station Subsystem

BTS Base Transceiver Station

CCS Common Channel Signaling

CDMA Code Division Multiple Access

CI Cell Identity

CN Core Network

CS Circuit Switched

D-AMPS Digital Advanced Mobile Phone System

DCS Digital Cellular System

DS Direct Sequence

ECSD Enhanced Circuit Switched Data

EDGE Enhanced Data rates for Global Evolution

EGPRS Enhanced General Packet Radio System

EIR Equipment Identity Register

ETSI European Telecommunications Standard Institute

FDMA Frequency Division Multiple Access

GCR Group Call Register

GERAN GSM EDGE Radio Access Network

GGSN Gateway GPRS Support Node

GMSK Gaussian Minimum Shift Keying


2

GoS Grade of Service

GPRS General Packet Radio Service

GSM Global System for Mobile Communications

HLR Home Location Register

HSCSD High Speed Circuit Switched Data

IMS Internet Multimedia Subsystem

IMS-2000 International Mobile Telecommunications 2000

IMSI International Mobile Subscriber Identity

IP Internet Protocol

IR Incremental Redundancy

ISP Internet Service Provider

ITU International Telecommunications Union

LA Link Adaptation

LA Location Area

LAN Local Area Network

MCS Modulation and Coding Scheme

MMI Man-machine interface

MMS Multimedia Message Service

MSC Mobile Switching Centre

MS Mobile Station

MSRN Mobile station roaming number

NACC Network assisted cell change

NSS Network Sub-system

OMC Operations and Maintenance Center

OSS Operations Sub-system

PSTN Public Switched Telephone Network

QoS Quality of Service

RAN Radio Access Network

RF Radio Frequency

RNC Radio Network Controller

RTP Real-time Protocol

SGSN Serving GPRS Support Node

SIR Signal-to-Interference Ratio


3

SMS Short Message Service

SS Spread Spectrum

TDD Time Division Duplex

TDMA Time Division Multiple Access

TRAU Transcoder / Rate Adapter Unit

TRX Transceiver

UDP User Datagram Protocol

UE User Equipment

UMTS Universal Mobile Telecommunications System

UTRAN UMTS Terrestrial Radio Access Network

VLR Visitor Location Register

WAP Wireless Application Protocol

WCDMA Wideband Code Division Multiple Access


4

1. Introduction

EDGE technology gives GSM the capacity to handle services for the third generation

of mobile networks. EDGE was developed to enable the wireless transmission of

large amounts of data at a higher speed than before.

EDGE will allow GSM operators to use existing GSM radio bands to offer IP-based

multimedia services and applications at theoretical maximum speeds of 384 kbps

with a bit-rate of 48 kbps per timeslot and up to 69.2 kbps per timeslot in good radio

conditions.

Implementing EDGE will be relatively easy and will require relatively small changes

to network hardware and software as it uses the same TDMA (Time Division

Multiple Access) frame structure, logic channel and 200 kHz carrier bandwidth as

today's GSM networks, which allows existing cell plans to remain intact.


5

2. Evolution from GSM to GPRS

network architecture

2.1 Cellular Platform Evolution

Cellular radio networks are generally divided into three generations.

Analogue cellular systems, such as Nordic Mobile Telephone (NMT), are considered

to be the first generation of cellular technologies.

The second generation is the present digital network generation which includes

systems like Global System for Mobile communications (GSM), Digital Cellular

System (DCS), Digital Advanced Mobile Phone System (D-AMPS), and Interim

Standard –95 (IS-95). The second generation includes also enhancements to GSM:

High Speed Circuit Switched Data (HSCSD), General Packet Radio Service (GPRS)

and Enhanced Data rates for GSM Evolution (EDGE). These enhancements are

called the generation 2G+ or 2,5.

According to International Telecommunications Union (ITU) specifications, the

third generation cellular networks will offer data transmission speeds up to 2Mbps.

Universal Mobile Telecommunications System (UMTS) is one of the mobile

communications systems being developed within the ITU framework known as

International Mobile Telecommunications IMT-2000.


6

GSM9.6kbps UMTS

2Mbps

HSCSD57.6kbps

GPRS115kbps

1999 2000 2001 2002 2003

EDGE384kbps

Figure 2.1. Evolution paths of GSM towards third generation networks

The different GSM evolution paths are shown in Figure 2.1. The data rates are the

maximum data rates theoretically provided by different systems. In reality,

maximum data rates are achieved only in very limited circumstances, if at all.

2.2 GSM

GSM, which was first introduced in 1991, is one of the leading digital cellular

systems. Originally a European standard for digital mobile telephony, GSM has

become the world's most widely used mobile system and it is in use in over 170

countries. Over 578 million subscribers use GSM in 400 different networks. GSM

networks operate on the 900 MHz and 1800 MHz waveband in Europe, Asia and

Australia, and on the MHz 1900 waveband in North America and in parts of Latin

America and Africa.

GSM is an open, standardized, non-proprietary system that is constantly evolving.

The growth of GSM continues unabated with more than 160 million new customers

in the last 12 months. Since 1997, the number of GSM subscribers has increased by

10 fold.


7

In addition to voice services, GSM simplifies data transmission to allow laptop and

palmtop computers to be connected to GSM phones. It provides integrated voice

mail, high-speed data, fax, paging and short message services capabilities, as well as

secure communications. GSM offers the best voice quality of any current digital

wireless standard. Furthermore, one of GSM strengths is the international roaming

capability. Roaming gives subscribers seamless connection to mobile networks. In

addition, GSM satellite roaming has extended service access to areas where

terrestrial coverage is not available. [Ericsson 2003][Gsmworld 2003]

2.2.1 Functioning of the GSM system

GSM uses narrowband TDMA as a multiple access method where eight

simultaneous calls can occupy the same radio frequency while using a full rate

speech codec (eight timeslots/200kHz). Using a halfrate speech codec, where two

users can share the same timeslot, can double the capacity. However, using the

halfrate codec has a deteriorating effect on speech quality.

Operators have multiple frequencies and thus GSM is in fact a combination of

TDMA and FDMA (frequency division multiple access) technologies. Each timeslot

is called a physical channel and can be used as a traffic channel and/or a control

(signaling) channel. Traffic and control channels are called logical channels.

In GSM, there are different frequency ranges for uplink (from the mobile station’s

transmitter to the base station’s receiver) and downlink (from the base station to the

mobile station) traffic. [Penttinen 2002]

2.2.2 Architecture of the GSM network

GSM network consists of network and switching sub-system (NSS), the base station

sub-system (BSS) and the operations sub-system (OSS), which controls the

functioning of the NSS and BSS (see Figure 2.2.). In addition to these three

elements, GSM network contains also other elements for for example billing and

voice mailbox services. [Penttinen 2002]


8

Figure 2.2. GSM network architecture

Base Station Sub-system

The main task of the base station sub-system is to connect the mobile stations to the

network and to switch the sub-system in briefly. The base station sub-system consists

of base transceiver stations (BTS) and base station controllers (BSC) and a

transcoder / rate adapter unit (TRAU). One BTS consists of equipment space, a

Transmitter/Receiver (TRX) with power supply, a combiner, a power splitter,

antenna cables, a mast, a scrambler, antennas and an antenna pre-amplifier.

One TRX transmits traffic on a single frequency if there is no synthetic frequency

hopping. As mentioned earlier, each frequency is divided into eight time slots, and if

full rate codec is used, each frequency can carry eight users at maximum. Not all the


9

available time slots can be used for speech or data transmission because some of

them are used in signaling (SDCCH channels). Usually one cell consists of from one

to six TRX, which means that it is possible that there is up to forty-five users per cell

depending on the number of SDCCH channels. The geometry of the cell could be

circular or conical. Circular cells are usually used in the countryside and in areas

where it is sensible to use omnidirectional antennas. On the other hand, conical cells

are used in areas where directivity is needed, for example freeway areas and cities.

Different cells are separated from each other with individual cell identities (CI).

Another essential element in the BSS system is the earlier mentioned BSC, the main

function of which is to control and manage the BSS and the radio channels. It

transfers signaling information to and from the mobile stations (MS) and manages

the handovers between the cells. In the GSM system this kind of assignments are

defined apart from the mobile switching center so that the MSC would have more

time to switch calls through. MSC connects the calls via the right BSC to the MS,

and BSC handles the events of the radio interface during connection. One BSC

controls several BTSs, and grouping them enables constituting location areas (LA).

It is also possible that base transceiver stations constituting the LA are in the area of

two different BSC’s.

Figure 2.3. The structure of the location areas

If the MS moves in a standby state from the location area to another it has to update

the location in the BSC. When the network knows the location area of each MS, it

enables that in case of incoming call the network has to call MS via the cells in the

certain location area only (paging). BSC controls the channel al-location and knows

all the time, which channels are in use and which are free in each cell in its area. On


10

the basis of the measurement reports sent by mobile stations the BSC also knows the

interference level of each channel. Normally BSC only reports for the MSC about

the performed channel changes but in the case of handover between two BSC’s the

MSC takes part of the channel change. The handover may occur e.g. because of the

field strength or the quality level. In the GSM system the decision of the handover is

made by BSC on the basis of the measurement reports of the MS and BTS.

In addition to all assignment mentioned above, the BSC also controls the parameters

of the radio interface. There are typically a lot of parameters in the radio interface

containing e.g. frequency hopping sequences of the TCHs and location updates.

Usually one MSC administers several BSCs, and several base transceiver stations are

typically connected at one BSC. The amounts depend on the capacity of the devices,

which changes between different manufacturers. The capacity also depends on the

capacity of the registers in the central system. [Mouly 1992] [Penttinen 1999] [Nokia

2000]

Network and Switching Sub-system

Network and Switching Sub-system (NSS) is composed of MSC’s and registers

related to it. Related registers are home location register (HLR), visitor location

register (VLR), equipment identity register (EIR), authentication center (AuC) and

group call register (GCR). NSS handles the connections between external network

and MS. NSS also handles internal connections in the GSM network. These

connections are links between two MSC’s and internal connections of MSC.

Specified interfaces are defined to the MSC because of the external connections for

both speech and data services. In these connections, fixed network numbering and

common channel signaling (CCS) are normally used. Because of data services,

matching functions against the external networks are defined.

MSC is the most important part of the NSS, and its main tasks are to connect,

maintain, and discharge connections in the area. The structure of the MSC is almost

identical to the structure of switching centers in the fixed networks. The structures of

these two different centers are so similar that the switching centers in the fixed

network sees the MSC as any digital transit exchange. Compared to the centers in the


11

fixed network, the MSC has special functions related to the e.g. mobility

management and connection protect.

One or more BSCs are connected to each MSC on the A-interface. There can also be

one or more connections to the centers of the external phone or ISDN networks. It is

specified that the MSC must be able to process a great amount of service requests

because the processor load of the MSC is multiple compared to the centers in the

fixed network. This is a consequence of the great amount of signaling concerning the

mobility management, handovers and other signaling functions characteristic of

GSM networks.

As mentioned above, there can be five registers connected to the MSC, but only two

of these five registers are necessary. Necessary registers are HLR and VLR and the

remaining three are only possible but not necessary.

In the HLR, there are a captured subscriber and billing information and the

supplementary services of the number. This information can be permanent or

changing. Permanent information includes mobile subscriber ISDN number

(MSISDN), international mobile subscriber identity (IMSI), encryption parameters

of the subscriber, and the type of the subscriber connection. Changing information,

on the other hand, includes elusiveness and routing information, and information of

the call transfer. Each subscriber is registered in only one HLR and the operator of

the home network makes this. In practice, the HLR is a computer equipped with a

large hard disk and this computer is connected to the MSC via the C-interface.

VLR contains the subscriber information of each MS outside the home network.

When the MS updates at the new MSC area, the VLR of that area requests the

subscriber information from the HLR and at the same time updates the location

information to the HLR. The subscriber information of the MS will be stored in the

VLR as long as the MS stays on the area of that VLR. In addition to HLR, the

subscriber information will always be found in one VLR. When the subscriber

moves in the network to the area of new VLR the subscriber information of the MS

will be deleted from the old VLR and removed to the new VLR. In practice the VLR

contains e.g. MSISDN and IMSI numbers, temporary mobile subscriber identity

(IMSI), mobile station roaming number (MSRN), and the in-formation about the


12

location area (LA). In addition, the VLR also contains the service specifications and

encryption parameter (encryption triplets). One VLR is always integrated on each

MSC, and for this reason the mobile switching center is often indicated with

acronym MSC/VLR. The VLR is connected to the MSC via the B interface. Because

of the MSC and VLR are always integrated, the B-interface is not specified. VLR is

connected to the HLR via the D-interface and the signaling to the other VLR

elements is carried out via the G-interface. [Mouly 1992][Penttinen 1999][Nokia

2000][ETSI 2000]

Operations Sub-system

Operations sub-system (OSS) is not specified very exactly. Consequently, a high

degree of freedom of choice is left for the network manufacturers. On the other

hand, the interfaces between OSS and other network elements are specified.

The OSS has several tasks, which demands connection with BSS or NSS. Most

important tasks of the operations sub-system are operation and maintenance of the

network and the management of the subscriber information and mobile stations.

There are one or more operations and maintenance center (OMC) in the OSS, in

which it is possible to install software of the network elements, enter parameters and

supervise the statuses of the network elements. OMC is in direct connection to the

MSC and the BSC, and to the BTSs via the Abis-interface of the BSC. OMC is

usable from the workstations via the man-machine interface (MMI).

OMC is connected to the other network elements with strictly specified Q3-interface

to enable compatibility between devices of different manufacturers. In practice, this

compatibility did never exist so there must be as many OMCs as there are devices

from different manufacturers in the network. [Mouly 1992] [Penttinen 1999] [Nokia

2000]

2.2.3 GSM mobile stations and SIM cards

Each GSM mobile station consists of the actual mobile equipment and the subscriber

identity module (SIM) card. The SIM card provides GSM system with the safety

functions and flexible adding and removing of services.


13

A mobile station can function on one or more frequencies (multiband) and in several

systems (multimode). [Penttinen 2002]

2.3 High Speed Circuit Switched Data

High Speed Circuit Switched Data (HSCSD) adds many improvements to the basic

GSM. The most substantial change is that multiple time slots can be allocated for

one connection (1-8). That is called a multislot technique. In addition, the new

channel coding increases the bit rate in one time slot from 9,6kbps to 14,4kbps.

Because of the circuit-switched nature of HSCSD, the access times to packet data

networks, e.g., the Internet and intranets are relatively high.

HSCSD is one major competitor to the GPRS competing the title of the fastest

mobile data service at the moment. HSCSD has been in use commercially since

1999.

RTSLs 9.6 kbps 14.4 kbps1 9.6kbps 14.4 kbps2 19.2 kbps 28.8 kbps3 28.8 kbps 43.2 kbps4 38.4 kbps 57.6 kbps

Table 2.1. Data rates achieved by HSCSD

HSCSD terminals are already on the market, but they do not support more than three

timeslots downlink and one uplink. They allow GSM high-speed data services for

faster web browsing or file transfers. High-speed functionality can be used when a

phone or a phone card with HSCSD capabilities is connected to a compatible

computer via an infrared (IR) connection, cable, PCMCIA or Bluetooth.

As mentioned earlier, the main idea of the HSCSD multislot technique is to use

several channels in data transmission. GSM specifications state that the maximum

amount of the channels used concurrently is eight but in the HSCSD Phase 1 use of

four channels only is supported at once. It is unlikely that networks and terminals

would support the functionality of using all eight timeslots for one connection in the

near future because of the complexity of the technique. In addition, a network can


14

rarely guarantee the simultaneous use of eight time slots for data for extended time

periods. [Penttinen 1999] [Nokia 2000] [Penttinen 2002]

2.4 General Packet Radio Service

General Packet Radio Service (GPRS) is a packet-switched enhancement of existing

GSM networks. It is developed to allow large amounts of data to be sent over

cellular networks at speeds three to four times greater than conventional GSM

systems.

Because GSM is the most widely used mobile system in the world, for most

operators GPRS is the easiest and most logical way of offering customers fast

simultaneous data services, such as multimedia messaging, gaming, entertainment,

and news.

GPRS has been implemented in many GSM networks since 2000. Currently 188

telecommunications operators have invested in GPRS technology, 78 networks are

already in commercial service. [Ericsson 2003][Penttinen 2002]

2.4.1 GPRS in general

GPRS is designed for the transmission of bursty data based on the Internet protocol

(IP). GPRS uses radio resources only when there is data to be sent or received, and is

thus well adapted to the very bursty nature of data applications. One main idea of the

GPRS system is that it will not necessarily consume any capacity of the circuit

switched functions in the radio path. This is possible because the GPRS system can

be parametered to use only the capacity that is left over from the circuit switched

calls and data transmission. GPRS takes the advantage of the over-capacity, which

would remain unused otherwise (see Figure 2.4.).


15

Figure 2.4. The usage of the on-demand GPRS channels

With GPRS data is handled as a series of packets that can be routed over several

paths through the network, rather than as a continuous bit-stream over a dedicated

dial-up connection.

GPRS splits information into packets, which are transmitted over any available

circuit. When there are no packets being sent by one phone, the circuits are made

available for data packets from other phones. This makes highly efficient use of

available network resources and enables the introduction of always-on mobile

communication. In second-generation mobile networks, calls are handled using

traditional circuit-switching technology. A dedicated circuit (timeslot) is allocated

between two points for the duration of a call. No other phone can use this circuit

during the call, regardless of whether any data is being transmitted.

GPRS has no dial-up time so it is always connected to the Internet. GPRS offers

session establishment times below one second. GPRS users have continuous access

to Mobile Internet services for as long as the phone is switched on.

In today’s GSM radio networks, individual time slots offer a data rate of 9.6 Kbps

(or 14.4 Kbps in some upgraded networks). GPRS uses the same time slots, but can

use several at the same time (multislot technique) enabling much higher data rates

without having to establish a dedicated connection.


16

The packet-switching capability provided by GPRS is fundamental to the

introduction of 3G mobile communications, which builds on IP-based

communications to deliver broadband mobile multimedia services. [Ericsson 2003]

[Penttinen 2002]

2.4.2 GPRS Architecture

The implementation of the packet switched data service causes several changes to the

existing GSM network.

New network elements will be needed when updating the GSM network for the

GPRS use. The BTSs and BSCs of the GSM network will need at least software and

possibly hardware updates depending on the device implementation and the version.

HLR can be updated in the software level. In addition to all this, an extension to the

GSM network is needed, in the form of whole new IP-based GPRS backbone

network. The backbone network can be built with the existing infrastructure using

for example ATM technique. Due to these changes several new interfaces will be

formed compared to the GSM network. [Penttinen 2002]

Two new network elements are required to handle GPRS applications: the Serving

GPRS Support Node (SGSN) and the Gateway GPRS Support Node (GGSN), see

Figure 2.5.

These new nodes are scalable so operators can plan the network expansion that is

most affordable and best fits subscriber’s needs. Operators can start by offering high-

speed packet data services using small nodes in selected areas cost-effectively, and

then add extra capacity, as it is needed. [Ericsson 2003]

The SGSN concentrates on serving and tracking the mobile. It provides packet

routing to and from a SGSN service area. The SGSN functions are authentication,

session management, Short Message Services (SMS), mobility management and

routing. Also, charging and security functions are performed.


17

Figure 2.5. Structure of GSM/GPRS network (Rantanen 2001)

The mobility management (MM) in the GPRS differs from the traditional GSM

system. In the GPRS there are three mobility management states named Idle,

Standby and Ready. Mobility management of the GPRS includes GPRS-attach and

GPRS-detach functions, security functions, routing and location updates, and the

activation and deactivation of the PDP context. GPRS differs significantly from the

GSM in the case of handovers because there are no handovers in the GPRS. The

GPRS system has only cell reselection, which is made autonomously by the mobile.

[Penttinen 2002]

The GGSN functions as an interface to and from external packet-switched networks.

The GGSN is connected with SGSNs via an IP-based GPRS backbone network. The

GGSN performs almost the same set of functions as the SGSN excluding SMS.

2.4.3 Logical Channels of GPRS

In addition to GSM channels, special packet channels to the GPRS are determined.

These channels are divided into physical and logical channels. The logical channels


18

are divided into signaling and traffic channels. Physical channels meant for packet

data are called packet data channels (PDCH). A PDCH is a timeslot on a GSM RF-

carrier and logical packet data channels are mapped on the physical channels

dedicated to packet data traffic. Packet data logical channels can be divided into four

groups. [Penttinen 2002]

2.4.4 GPRS coding schemes

Channel coding is a technique used to protect the transmitted data packets against

errors. Four channel coding schemes are defined on GPRS standards for packet data

traffic channels. These coding schemes are marked as CS-1…CS-4. CS-1 has highest

error correction and lowest data throughput and the channel coding technique it uses

is the same as in the SDCCH of the GSM (ETSI Specification GSM 05.04). Coding

schemes 2 and 3 are punctured versions from CS-1. In CS-4 channel coding is not

used at all.

The more efficient channel coding used, the smaller is the proportion of the payload

in the emission. Thus, higher data rates are achieved by reducing or removing the

error correction bits. Table 2.2. presents the most important parameters for different

coding schemes.

Class Code rate Payload Data rate (kbps)

CS-1 1/2 181 9,05CS-2 ~2/3 268 13,4CS-3 ~3/4 312 15,6CS-4 1 428 21,4

Table 2.2. Most important parameters for GPRS coding schemes.

Channel coding schemes have a straight correlation to the C/I ratio in the way that

the lower the channel coding scheme used, the better the C/I ratio must be. On

account of this, all channel coding schemes have their own dissentient coverage

areas. The coverage area achieved with CS-1 is the same class as the coverage areas

in the traditional GSM systems with same coding scheme. The weaker the coding

scheme used, the smaller the achieved coverage area is. In practice, this means


19

highest data rates near by the base station and lowest near the borders of the

coverage areas (see Figure 2.6.). [Penttinen 2002]

Figure 2.6. Principle of the coverage areas achieved with different coding schemes

2.4.5 GPRS terminals

GPRS terminals (GPRS MS) are divided into three classes according to their

functionality:

Class A is the most demanding class of GPRS terminals. A terminal of this class is

able to establish simultaneous connections both with circuit switched (CS) and

packet switched (PS) side of the network. Class A terminals are not available on the

market.

Class B is able to select automatically either circuit switched or packet switched

connection but only one can be active at the time. So if the MS has both CS and PS

connection formed another service is in wait condition.

Class C terminals cannot be attached to both services at the same time and the

selection of the operation mode must be done manually. This class includes a special

case, terminals supporting only packet switched services. [Penttinen 2002]


20

2.5 Evolution of GSM Data Services towards EDGE

Voice services are still the most important services provided by mobile

communications networks. However, the earlier presented enhancements to ordinary

GSM technology bring new possibilities for various data services. At the same time,

the importance of other than voice services grows rapidly.

The development of mobile data services follows the evolution path of cellular

technologies. The first generation analogue cellular systems offered extremely slow

and unreliable data connections and identification methods were not well developed.

The second-generation digital cellular systems made an improvement to the data

services and data rates. In addition, Subscriber Identification Module (SIM) cards in

GSM phones improved security and enabled, for example, safe bank connections and

using of cellular phones for money transactions.

The Short Message Service (SMS) provides guaranteed delivery of small data

packets even if the phone is switched off when the message is first sent.

Now, second-generation services offer higher bit rates and packet-switched

connections. The development path advances towards UMTS and third generation

services that offer the ground for many high-speed services. In addition, wearable

computers and totally computerized homes can be a part of everyday life after a few

years. Wireless Local Area Network (WLAN) products and other possible wireless

network applications can have a remarkable role in parallel with advanced cellular

network services.

2.5.1 Short Message Service

Short Message Service (SMS) is included in GSM design to fulfill the customer need

for speaking and paging from one single terminal [Mouly 92]. Short Message

Service (SMS) can be seen as the first packet-oriented data transmission service

implemented in cellular radio networks. According to ETSI standard, the actual SMS

message can be up to 160 characters long. However, with some mobile terminals

also longer messages can be written, but the text is divided in to several single

messages before it is sent to the receiver.


21

SMS messages have been the fastest growing area of mobile telecommunications

during the last few years. The current SMS in GSM systems offers people the

opportunity to send and receive simple pictures and messages and it also provides a

possibility to access different information and entertainment sources such as weather

reports, bus timetables and jokes. Also paying with SMS messages is possible, e.g.

tram tickets. SMS transfer will be implemented also in GPRS, EDGE and UMTS.

2.5.2 Wireless Access Protocol

The Wireless Application Protocol (WAP) is an advanced intelligent messaging

service for mobile terminals. The WAP specification is published by WAP Forum,

which creates license-free standards for the entire industry to use or develop

products.

WAP was introduced on markets in 1999 and in 2001 there were over 18 million

WAP users and over 50 million WAP-enabled handsets shipped worldwide. In

addition to information, messaging and entertainment services, WAP is used for

transactions demanding security: banking, finance, and M-commerce. However,

WAP has not been as successful as it was meant to be. Currently WAP functions

over GSM and GPRS, in the future it can be utilized with EDGE and UMTS

terminals.

WAP is based on The Wireless Markup Language (WML), which complies with

XML standards. It is used in order to provide, for example, Web pages to the mobile

devices. The language is designed to enable effective applications within the

constraints of handheld devices. WML provides a smaller, telephony aware, set of

markup tags. This capability makes it more appropriate to implement within

handheld devices than Hypertext Markup Language (HTML). This means that

HTML coded pages have to be converted into WML code before they can be used in

the WAP mode. [WAPForum 2003]

2.5.3 Multimedia Message Service

The multimedia message service (MMS) has become a significant issue for mobile

operators future growth strategies. MMS is expected to be the most important service


22

for operators, content providers and service providers since MMS will provide them

with a new source of revenue now and in the 3G markets.

The key to MMS is to maintain the fundamental features that have made SMS a

success story, while offering consumers a more versatile and personal experience.

MMS will enable consumers to send and receive multimedia messages between

mobile terminals as well as between terminals and content servers. MMS messages

combine image, sound and text, and even animation and video. The camera phones,

which are currently available, do not produce pictures, which are bigger than

100Kbytes.

Now, mobile operators have started to push MMS services seriously. At the start of

2002, only a single operator, Norway’s Telenor, had launched MMS-based picture

messaging. By November 2002, over 60 operators worldwide were offering picture

messaging.

One of the key factors that will determine the answer to whether or not large

numbers of consumers will take picture messaging part of their lifestyle is service

pricing. MMS-compatible phones are expensive and users will only be persuaded to

buy these camera phones if they can afford to use them. The widespread usage of

SMS text messaging has been enabled by service pricing which is both easy to

understand and fairly cheap. [Nokia 2003][Ovum 2003]

2.5.4 HSCSD and GPRS enabled services and data rates in practice

Today, HSCSD and GPRS connections are mainly used for accessing email, getting

information from the Internet, web surfing in general, entertainment (e.g.

downloading video clips, music, etc.), banking and shopping. The data rates

achieved by using HSCSD and GPRS with terminals available on the market are

quite similar.

The current terminals and networks do not support much over 40 kbps data rates in

practice – this means slower and more unstable connections than ordinary fixed-line

modems can offer.

HSCSD connections are more stable compared with GPRS connections because of

the circuit-switched nature of HSCSD. Even though, it is not assured, that a HSCSD


23

user gets all the three timeslots his terminal can carry for the downlink traffic. In

many cases, circuit-switched speech is prioritized over HSCSD and GPRS traffic and

thus timeslots first allocated for data are allocated for speech on the go. In the

current GPRS networks only CS-1 and CS-2 coding schemes are used. The average

C/I from tested networks leads to 11,5 kbps/timeslot. Therefore in practice, the

networks offer “best effort” service quality and slow connections.

Giving data users enough capacity is also a pricing issue. Adding capacity to the

network is always an extra investment. Operators in Finland have different pricing

strategies and thus also their network parameters for GPRS traffic differ from each

other.


24

3. EDGE technical fundamentals

Enhanced Data rates for GSM Evolution (EDGE) is a major enhancement to

GSM/GPRS data rates and it improves the GSM air-interface performance

significantly. EDGE offers improved data rate through optimized modulation (8-

PSK) and it introduces a large number of channel coding schemes along with

Incremental Redundancy (IR), Link Adaptation (LA) enhancements and in the near

future adaptive multirate AMR.

The new modulation and the possibility to adapt the transmission rate to channel

quality are the core of the EDGE concept. Introducing EDGE in a GSM network

does not imply changes in the basic architecture. In any case, modifications of the

Mobile Station (MS), Base Station (BTS) and Base Station Controller (BSC) are

needed, which means, among other things, software and hardware upgrades in

circuit- and packet-switched parts of the network.

EDGE offers both circuit- and packet-switched connections depending on the

platform it is implemented in. The scope of the EDGE phase 1 standard is to increase

GPRS bit rate, improve GPRS link quality control (EGPRS) and to offer high

circuit-switched data rate with fewer timeslots and fast power control (ESCD). The

scope of the EDGE phase 2 includes supporting real-time services over EGPRS.

GSM networks have already offered advanced data services from single SMS and

circuit-switched 9,6 kbps data services to 64kbps HSCSD and 160 kbps (theoretical


25

speed) GPRS. With EDGE, the data rate offered by the original HSCSD or GPRS

networks can triple. [Halonen et al 2002]

3.1 8-PSK modulation in GSM/EDGE standard

EDGE is specified to reuse the channel structure, channel width, channel coding and

the existing mechanisms and functionality of GSM, HSCSD and GPRS. The

enhancement behind tripling the data rates is the introduction of the new modulation

type.

The modulation type that is used in GSM is the Gaussian minimum shift keying

(GMSK), which is a kind of phase modulation. EDGE introduces the octagonal

phase shift keying (8-PSK) modulation in addition to the existing GMSK, see Figure

3.1.

(0,0,1)

(1,0,1)

(d(3k),d(3k+1),d(3k+2))=

(0,0,0) (0,1,0)

(0,1,1)

(1,1,1)

(1,1,0)

(1,0,0)

Figure 3.1. 8-PSK signal constellation principle

The number of symbols sent within a certain period of time, the symbol rate,

remains the same as for GMSK but an 8-PSK signal is able to carry three bits instead

of one. The total data rate is therefore increased threefold.

An 8-PSK modulated signal is more sensitive to errors and thus the highest data rates

can only be achieved with limited coverage. GMSK modulation is more efficient

under very poor radio conditions and therefore EDGE coding schemes are a mixture

of both GMSK and 8-PSK. [Ericsson 2002][Halonen et al 2002]


26

3.2 Enhanced general packet radio service (EGPRS)

Enhanced general packet radio service (EGPRS) is build on top of GPRS.

Four different coding schemes are defined for GPRS (CS-1 to CS-4). Each has

different amounts of error-correcting coding that is optimized for different radio

environments. For EGPRS nine modulation and coding schemes (MCS) are

introduced. Classes MCS-1 – MCS-4 use the basic GSM 0.3 GMSK modulation,

whereas classes MCS-5 – MCS-9 use the new 8-PSK modulation. Table 3.1. shows

EGPRS modulation and coding schemes along with their maximum throughputs.

Table 3.1. EGPRS modulation and coding schemes

Another improvement that has been made to EGPRS standard is the ability to

retransmit a packet that has not been decoded properly with a more robust coding

scheme, whereas for GPRS re-segmentation is not possible. In GPRS once packets

have been sent, they must be retransmitted using the original coding scheme even if

the radio environment has changed.

3.2.1 Link adaptation

EGPRS uses automatic link adaptation (LA). LA is used to select the best MCS for

the radio link conditions. LA uses the radio link quality measured either by the

mobile station or by the base station to select the most appropriate modulation and

coding scheme for transmission of packets. Each modulation and channel coding

class is optimized for a certain range of C/I (interference) values, outside of which


27

the data rate no longer increases together with the C/I value, but saturates. LA

algorithms compare the estimated channel quality to threshold values and that leads

to optimized throughput. In reality, the link adaptation may not be close to the ideal

situation where the maximum data rate is (as a function of the C/I curve) achieved

by switching channel coding class “on the go”. [Ericsson 2002][Halonen et al 2002]

3.2.2 Incremental redundancy

Another way to choose the optimal channel coding class is to use the incremental

redundancy technique (IR). Incremental redundancy initially uses a coding scheme

with very little error protection (such as MCS-9) and without consideration for the

actual radio link quality. When information is received incorrectly, additional coding

is transmitted and the resent information is soft combined in the receiver with the

previously received information. IR adjusts the code rate of the transmission to true

channel conditions with incremental transmissions of the redundant information until

the decoding is successful. For the mobile stations, incremental redundancy support

is mandatory in the standard. The information about the radio link is not necessarily

to support incremental redundancy. IR gives additional 2-3dB to the radio link.

[Ericsson 2002][Halonen et al 2002]

3.3 Enhanced circuit switched data (ECSD)

Enhanced circuit switched data (ECSD) is based on the current HSCSD is GSM

networks. The ECSD architecture is mainly based on HSCSD transmission and

signaling.

ECSD does not increase the maximum 64 kbps data rate of HSCSD but it makes the

network more efficient: the same data rates can be achieved with allocation of fewer

timeslots and simpler MS implementation.

Circuit-switched data connections up to 64kbps are sufficient for providing various

transparent and non-transparent services, e.g. interworking with audio modems and

ISDN at various data rates and various video based services ranging from still image

transfer to videoconferencing services.


28

In the future, Enhanced Adaptive Multi Rate codec (EAMR) enables the transfer of

high-quality speech and music. The same restrictions apply to EAMR connections as

apply to ECSD, as EAMR is also circuit-switched. [Halonen et al 2002][Penttinen

2002]

3.4 EDGE evolution towards GERAN Rel´5

GSM/EDGE radio access network (GERAN) Rel´5 includes a definition of

enhancements to the GPRS radio link interface and will provide support for

conversational and streaming service classes as defined for WCDMA. With the

adoption of the UMTS Iu interface and the UMTS quality of service (QoS)

architecture in Rel´5, GERAN and UTRAN can be efficiently integrated under a

single UMTS multi-radio network. In addition, GERAN will include performance

enhancements for existing services.

IP Network

HLR

MSC/VLR

SGSN

RNC

RNCBTS

PSTN

GGSN

UTRAN

Network Subsystem

GPRS-backbone

BTS

EDGE BS

BSCBTS

GPRS/EDGE Radio Network

Core NetworkUMTS Radio Network

Figure 3.2. A simplified model of the combined GSM GPRS/EDGE and UMTS

network (Rantanen 2001)

In general, the goals and impacts of GERAN Rel’5 specification are to enable

GERAN to the same 3G CN (core network) as UTRAN creating first steps towards

efficient resource optimizations in multi-radio networks, and to enable GERAN to


29

provide the same set of services as UTRAN, making the radio technology invisible

to the end-user, while allowing operators to efficiently manage the available

spectrum. The existing GERAN radio protocols need to undergo significant

modifications, and this will increase the complexity of radio interface protocols. In

addition, standardization of the GERAN Rel´5 should support a true multi-vendor

environment and GSM/EDGE radio access should be backwards compatible, i.e.

support of services for pre-Rel’5 terminals must be ensured. [Ericsson 2002]

[Halonen et al 2002]

3.4.1 GERAN Rel´5 features

In the 3rd Generation Partnership Project (3GPP) Rel´5 overall, the most significant

new functionality is the Internet multimedia subsystem (IMS). From the GERAN

perspective, the support for the IMS services implies introduction of the Iu interface,

and definition of the header adaptation mechanism for the real-time protocol (RTP),

user datagram protocol (UDP), and Internet protocol (IP) traffic. Rel´5 includes also

major enhancements for speech: wideband AMR speech for enhanced speech

quality, half-rate 8-PSK speech for improved speech capacity, and fast power control

for speech. In addition to the abovementioned enhancements, Rel’5 implies location

service enhancements for Gb and Iu interfaces and inter-BSC and BSC/RNC

network assisted cell change (NACC). [Halonen et al 2002]

3.4.2 GERAN Rel´5 system architecture

To connect to the WCDMA/GPRS core networks, GERAN will use the Iu interface,

as shown in Figure 3.3. The Iu interface connects to the circuit-switched domain (Iu-

cs) and to the packet-switched domain of the core network (Iu-ps). GERAN also

connects to the second-generation core network nodes by using the A and Gb

interfaces. These interfaces remain intact in Rel’5 to support Rel’99 terminals. Iu-ps

interface is not used for Rel´99 terminals because the functional split between the

radio access network and the core network differ substantially between Iu and A/Gb.


30

BSS

RNC

Um

GERAN

UTRAN

BSC

BTS

BTS

GSM/WCDMA Core Network

MS

MS

Iur-g

A

Gb

Iu

Iur-g

BSS

RNC

Um

GERAN

UTRAN

BSC

BTS

BTS

GSM/WCDMA Core NetworkGSM/WCDMA Core Network

MS

MS

Iur-g

A

Gb

Iu

Iur-g

Figure 3.3. GERAN architecture in Rel’5

The radio interface Um between GERAN and the mobile station is based on the Rel

´99 radio interface link specifications. However, several modifications are needed on

radio link protocol layers in order to provide adequate radio bearers for real-time

services. These modifications imply to support for cell reselection for packet-

switched domain, separation of user and control planes, and transparent modes in the

radio link protocol layers. [Ericsson 2002][Halonen et al 2002]

3.5 Modifications to the GSM network imposed by EDGE

The implementation of GSM EDGE requires basically only TRX change to EDGE

TRXs in the GSM base stations and software updates to GSM BSC and GPRS IP-

backbone. A bigger investment would most probably be the upgrade of Abis

interface from 16 kbit/timeslot connection to EDGE capable 64 kbit/timeslot

connection. [Rantanen 2001]

The impact of EGPRS on the existing GSM/GPRS network is limited to the base

station system due to the minor differences between GPRS and EGPRS. A new

transceiver unit capable of handling EDGE modulation as well as new software that


31

enables the new protocol for packets over the radio interface in both the base station

and base station controller. The core network remains intact. [Ericsson 2002]

A-bisA-bis MSC

GnGn

GGSN

BSC

AA

2G SGSN

BTS

BTS

OSS

GSM/EDGE

IuIu

Figure 3.4. EDGE implementation [Auramo 2002]

In Ericsson case, EDGE is compatible with recent equipment. If an operator has an

Ericsson RBS 2000 macro base station from 1995 or later, it is easy to take on

EDGE. Some additional hardware using plug-in transceivers, and new software that

can be installed remotely is all that is needed for operators to start offering high-

quality Mobile Internet services over their existing infrastructure. [Ericsson 2003] In

Figure 3.4, the elements for EDGE implementation are shown.

3.5.1 EDGE radio network planning compared with GSM/GPRS planning

If we think the implementation of EDGE of the radio network planning perspective,

the same principles as in the GSM/GPRS network planning apply. As in GPRS,

EDGE performance is dependent on the achievable C/I (and RXlev) in the network.

The most effective means to gain high performance in good radio conditions is to

come up with a optimized frequency plan. Frequency plan optimization can make a

significant difference for the achievable throughput.


32

Propagationestimations

CoverageAnalysis

Interference matrix

•co-channel•adjacent channel

Frequency plan

Separationconstraints

TRX requirements

Figure 3.5. EDGE network planning

EDGE deployment doesn’t bring dramatic changes to radio network planning with

GPRS. Main concerns are the allocation of capacity and steering of traffic to wanted

layer/cell/TRX. EDGE network planning process is shown in Figure 3.5. Changes to

transmission capacity will be needed, if larger scale EDGE deployment per cell/area

is done.

The easiest way to implement EDGE from the network planning point of view is the

TRX replacing strategy, where new frequency plan is not mandatory. The replacing

can be done for every 1-3rd site to achieve coverage and EDGE services e.g.

hotspots or rural area can be selected for EDGE, but with limited amount of data

throughput.

Higher data amounts with EDGE can be offered if it is implemented by bringing an

additional EDGE TRX dedicated to data usage to (some of) the cells in the network

and/or by reserving more timeslots for the use of EDGE data users from the TRXs.

However, that leads to decrease in the GoS experienced by the speech users. In real

life these actions are not always possible to perform and they will require

significantly more implementation and planning work.

In order to utilise EDGE performance in full, a totally new frequency plan and

possibly new GSM cell structure are required.


33

4. Vendors EDGE strategies

Due to the delays in UMTS implementation compared to the early predictions, most

vendors have taken EDGE back to the table. EDGE as a technology was firstly

developed to support the GSM evolution towards 3G in especially US markets. As

UMTS hype began in 1999/2000 EDGE was put to less priority among most of the

vendors, because UMTS implementation seemed to happen so fast and with large

scale that it was natural to shift all focus to support this.

After the enormous UMTS license fees most of the operators’ capabilities to invest

in to the networks were decreased significantly. At the same time making the UMTS

technology work needed more work from the vendors than anticipated, which

together with the operators decreased investment capabilities caused the delays in

UMTS implementation. EDGE was again an issue, because the needed investments

on it are a fraction of those needed for UMTS and the end user performance is quite

close to UMTS in the beginning. Of course the capacity offered by UMTS is

enormous compared to EDGE, but as there is no pressure on the capacity side for the

operators (because the data traffic haven’t proved to increase still) UMTS capacity is

not really needed yet.

When the US operators, for example AT&T, started to implement EDGE capable

HW as they decided to go for the GSM evolution towards 3G rather than IS-95

based, there was suddenly a need for the operators to start making EDGE terminals

as well. As the biggest reason for the US operators to choose GSM based system is


34

the roaming traffic gained globally, there is a clear need to have EDGE happening in

the Europe and Asia as well. This was the reason why EDGE marketing started again

with full steam for European operators as well by the biggest vendors.

The biggest two vendors, Nokia and Ericsson, are the most active with EDGE

marketing. Nokia has been clearly the market maker in Europe and Asia Pacific for

EDGE. It’s of course in interests of all the vendors to make EDGE a success, but due

to its strong position in terminal market, Nokia is in better position to drive the

market than it’s competitors. Ericsson has clearly taken a follower role in EDGE

market, focusing clearly on driving the UMTS market. This is also partly due to the

difficult financial situation of Ericsson currently, where the ability to take risks in

new market areas is limited. When the two vendors are compared from the point of

view of needed increments to the legacy network infrastructure (to make EDGE

possible), Ericsson is in stronger position than Nokia with better applicability of the

older infrastructure in the field.

All the other noticeable GSM network vendors (Siemens, Alcatel, Motorola, Nortel)

have taken a reactive role with EDGE, waiting for the market to start up. Outside

USA there has been little marketing done for EDGE by these vendors and they are

clearly waiting and seeing whether the big vendors (mainly Nokia) can have the

market created for EDGE and then jumping on board. Of course they all have EDGE

infrastructure and terminals as well in their road-maps, but they are not put into

number one priority and committed on.

The problem with EDGE is for a network vendor that it has been earlier positioned

as 3G technology thus competing with the UMTS market. So, as the potential UMTS

market is clearly bigger than EDGE, it has been decided by most vendors not to

drive EDGE strongly towards their customers. This could have an negative effect on

the UMTS sales. The answer is to position EDGE more as a enhancement to existing

GPRS networks and to co-exist with UMTS.

Currently as first Nokia and then Sony-Ericsson have committed to bring EDGE

terminals to Europe-Asia GSM bands as well, it seems that the market is clearly

starting. The availability of terminals and thus necessary penetration is in vital role

in possible EDGE success.


35


36

5. Services enabled by EDGE

The Release 99 EDGE implementation does not offer significant new possibilities

for services compared with the current HSCSD and GPRS networks.

As mentioned before, circuit-switched data connections up to 64kbps are sufficient

for providing various transparent and non-transparent services, e.g. interworking

with audio modems and ISDN at various data rates and various video based services

ranging from still image transfer to videoconferencing services. Packet-switched

connections are optimal for bursty data traffic, e.g. web browsing and email.

In Rel’5 UMTS 3GPP traffic classes are enabled in EDGE and thus 3G services

delivery across all frequency bands and bearers becomes possible. Handovers across

GSM/EDGE/WCDMA are enabled from the start. However, there are still

uncertainties in standardization of Rel´5 and the Iu interface.

When compared to GPRS phase 1 QoS classification, the QoS grouping of UMTS

release 99 takes into account the applications that will become available through the

increased data rates of UMTS and EDGE. The main distinguishing factor between

the traffic classes is the sensitiveness of applications, as presented in Table 5.1.

[Rantanen 2001]


37

Table 5.1. QoS classes for UMTS and EDGE

Conversational and streaming classes are mainly intended to be used to carry real-

time traffic flows. The main difference between them is the delay sensitiveness of

the classes. Interactive class and background class are mainly meant to be used by

the Internet type applications e.g. web browsing and e-mail. Due to looser delay

requirements compared to conversational and streaming classes, both provide better

error recovery by means of retransmission.

Datarate

Con

vers

atio

nal

Bac

kgro

und

0 8 16 48 128 473 2048

Con

vers

atio

nal

Inte

ract

ive

Con

vers

atio

nal

Strea

min

gCon

vers

atio

nal Voice

Corporatesolutions

InfotainmentVoice

FAXCollaborative working

Communication

Transactionservices

Advertising

Audio clip downl.

Video clip downl.Short

Messaging

Corporate Data Access

WEB Browsing

WAP Applications

E-mail

Video streaming

Multimedia Messaging

Video telephony

Audio streaming

Gaming

WCDMAWCDMA

EGPRSEGPRS

GPRSGPRS

Figure 5.1.Service QoS Requirements for Bearers, Data rates and services 2003

[Auramo 2002]

The main difference between interactive and background class is that interactive

class traffic will have higher priority in scheduling than the background class traffic.


38

This means that background applications may use transmission resources only when

other applications do not need them. [Halonen et al 2002]

Although the conversational class is specified in the QoS classes of UMTS release

99, the most delay-critical applications such as speech and video telephony will be

carried on circuit-switched bearers in the first phase of the third generation mobile

networks. Later it will be possible to support delay-critical services as packet data

with QoS functions. Different QoS requirements of services are shown in Figure 5.1.


39

6. Terminal availability

The first terminals for EDGE will be on the market by 2H2003 by Nokia and

Motorola. These are aimed for the US market, but will have also European GSM

frequencies available. For example the Nokia 6200 will have GSM 1900/1800/800

frequencies and support for MCS1-9. With the same timetable Nokia will also

introduce an EDGE terminal with GSM 900 frequency available.

The Motorola t725 will have also both European frequencies imbedded and supports

MCS1-9. The t725 will be available by 2H2003.

Sony-Ericsson has also said that it will bring EDGE capable terminals for both

European and US markets in the second half of 2003. An interesting product will be

also the PC-card with GPRS/EDGE from Sony-Ericsson, which will be available

also in the second half of 2003.


40

Nokia 6200

Motorola t725

The most important fact when EDGE terminals are considered is the information

given by Nokia, that it will include EDGE to all of its GPRS terminals that are

introduced after 6/2003. Nokia has also said that it will have EDGE included in all

of its terminal categories from the beginning of 2004. This is extremely good news

for EDGE, since it almost positively ensures that the terminal penetration will start

to develop and that other vendors will join in manufacturing EDGE terminals. It also

gives operators a positive signal to include EDGE in their network evolution strategy

as a realistic option.


41

7. Investment costs and revenues

caused by EDGE

EDGE is a further development of the GSM/GPRS networks and thus it can be

integrated in the already established GSM/GPRS networks with relatively low

investment. Beside from the hardware and software upgrades, it only affects the

network by increasing capacity and data rates. The cost of operation will not

increase. Operators can deploy EDGE using the existing GSM spectrum. EDGE has

higher spectral efficiency than GSM/GPRS and thus there is free capacity for

carrying more data and voice traffic and for serving more subscribers.

The size of the investments on EDGE depends on the operator. The worst case is that

large part of the network infrastructure is old enough to not support easy EDGE

implementation. The capability of the infrastructure depends of the network vendor.

In some cases the base stations must be fully replaced by newer ones before EDGE

can be implemented. In typical cases only EDGE capable TRXs and BSS software

must be implemented along with changes to Abis interface capacity. If the network

infrastructure is new enough, the base stations can be already equipped with EDGE

capable TRXs. Then only software and enhanced transmission capacity must be

implemented and thus costs can be kept quite minimal.

In typical cases of operators’ network evolution the changes are made concurrently

as much as possible. This means that for example if the network infrastructure is old


42

enough to require new base stations if EDGE is considered, can the changes planned

so that UMTS/EDGE capable base stations are introduced at the same time. This

lowers the needed investments (and the operational expenses) for EDGE, which

would separately be quite enormous. Similarly EDGE capable infrastructure can be

moved to replace older infrastructure in the needed areas if available. This also

makes the investments lower. Depending on the situation of the operator, the

following costs are related to EDGE implementation:

EDGE capable GSM/GPRS base stations

EDGE capable TRXs

EDGE capable BSS software

Enhancements to Abis capacity

NMS/OMC changes

Possible upgrades to GPRS core network

Network planning costs (site configuration planning, frequency planning etc.)

Operational costs of implementation

Since the investments needed for EDGE are highly dependant on operators network,

strategy and cost structure and network vendor’s capability and pricing towards a

specific operator, it’s quite impossible to give generic cases of the needed total

investments. It can however be said that they are a fraction of that needed for

UMTS. EDGE can be implemented to every third site for example, so it enables lots

of different capacity/coverage strategies, which can be used to optimize the costs

involved. Some practical cases have shown that the pricing for EDGE TRXs and

BSS software is quite similar or only a bit higher to that of GPRS equipment. This

can also be due to the fact that the vendors need the reference networks up and

running, which usually means that the margins for the sales are kept lower than

usual.


43

As the EDGE capable terminals reach a feasible penetration percentage, the data

traffic is more economical to be served with EDGE rather than GPRS, this is

because the capacity provided by EDGE is almost 3-fold compared to GPRS, with

relatively small investments needed. This fact also allows more capacity for speech

service, taken that the network is parameterized accordingly. This enables greater

revenues for an operator. Also the higher throughput offered by EDGE for the users

can be priced higher than the conventional GPRS. Later if the GERAN offers same

QoS functions as UTMS it will create even more possibilities to generate revenue

from the users.


44

8. EDGE investment strategies

Different second-generation systems have different evolution paths towards third

generation IMT-2000 services. Network operators that are not granted UMTS

licenses can implement EDGE to offer IMT-2000 alike services. However, an

operator with UMTS license may still deploy EDGE to create a wireless data market

before third generation CDMA systems are launched or use EDGE in areas where

there is no UMTS service.

8.1 GSM operators without 3G licenses

For those operators without 3G license, EDGE offers a pretty straightforward

business case as a stepping stone before UMTS. Of course an operator can choose

not to go for UMTS at all and offer high data rate services with GPRS/EDGE

network. This type of stepping stone approach is currently used by some of the US

operators (such as AT&T), which have not decided the UMTS bandwidth and are

implementing GSM-EDGE networks as this is written. In Europe most of the

countries have already granted their licenses so there are very few left to implement

this strategy. In Asia, there is a technology standard war going on between

CDMA2000, WCDMA and TD-SCDMA. In this area EDGE is clearly seen as a less

attractive option.

EDGE as a stepping stone to UMTS can be seen only feasible in markets with no

strong UMTS commitment: No licenses yet awarded or they are very inexpensive or


45

no present market push for 3G or if the GSM business is still developing. Besides

US, this kind of situation exists e.g. in East-Europe. Although if the first UMTS

launches (e.g. Hutchison UK, Italy) are successes and UMTS terminals come faster

to market and are more strongly subsided than the EDGE terminals, then the window

for EDGE feasible will become smaller rapidly for this type of strategy. The only

clear situation is in the US, because they will not have UMTS licenses awarded for

some time.

If this type of strategy is chosen by the operator, all the services can/will have to be

planned on top of GSM-EDGE. This means that if for example streaming services

are to be offered must EDGE standardization support this. In the first releases of

EDGE the QoS will be similar to GPRS, which doesn’t enable similar services as

can be offered by the UMTS.

EDGE as a stepping stone can require EDGE to be implemented over the network,

which of course will make the investments bigger as well. If the network is built

directly to support EDGE then the investments can be made smaller. Of course

EDGE can be implemented only to e.g. cities and GPRS elsewhere, which lowers the

incremental costs.

8.2 GSM operators with 3G license

EDGE and WCDMA can also be complementary 3G technologies that together will

sustain the operator’s need for nation-wide mobile data and speech capacity during

the expected traffic boom. They are both IMT-2000 capable radio access

technologies in different frequency bands. They can both provide 3G services for the

end-user, accessing a common core network, given that the Iu-interface is

standardized to be supported by the GERAN as well.

Possible business cases for the operator with a UMTS license are for example the

following:

EDGE used as a complementary solution, to different coverage areas. In this

case EDGE is implemented e.g. to more rural areas and UMTS to the cities and sub-

urbans and main roads. This case makes it possible to offer high-data rate services

also in the areas where UMTS coverage is not available. The investments of EDGE


46

implementation can be made lower by guaranteeing that the infrastructure in the

“EDGE areas” is the most EDGE capable one, if possible by swapping.

This case requires though that there will be multi-mode terminals available quite

rapidly (i.e. UMTS/EDGE/GPRS), because otherwise the experienced throughput

can be higher in the rural areas than in the cities, which is not very feasible.

EDGE and UMTS co-exist, for different user segments. In this case EDGE and

UMTS are implemented to same coverage areas, but used to serve different user

segments. For example EDGE could be used to offer robust data for corporate access

and UMTS to offer fancy 3G services, such as video streaming etc.

This case will require quite large investments, because EDGE is to be implemented

over the network or at least to most locations. This case would also require that there

would not be multi-mode terminals widely available very rapidly, so the

segmentation could be made more easily through terminals.


47

9. Future role of EDGE

In Figure 9.1, the current situation of EDGE globally is depicted.

Latin America:Will eventually

follow US.

US+Canada:EDGE roll-outs on the way and EDGE will be

deployed during 2003

APAC:Market follows global

trends. “Ongoing technology standard

war”. Also public commitments to EDGE

China:Political

commitments to every

technology. No rush to

3G. No public EDGE

commitments yet

Europe:WCDMA

technology commitment.

Strong need for delaying

UMTS roll-outs

Growing interest

towards EDGE, but no public commitments

yet.

Global EDGE Status

Figure 9.1. Global situation of EDGE

As can be seen from Figure 9.1, the only market areas in which public commitments

to EDGE have been made is the US and Asia Pacific. In Europe, which is the key

market area for a larger scale EDGE success, has not yet seen any public EDGE

commitments from the operators. Although the biggest vendors claim to have

multiple contracts for EDGE deliveries, no of them are public yet in Europe.


48

The future role of EDGE depends very highly on the following:

The success of EDGE in USA

The availability of EDGE terminals

The early success of UMTS

Subsidisation of UMTS terminals

The completeness of EDGE standardization

The biggest guarantee that EDGE will become one of the steps of GSM network

evolution, which be implemented as well, is the availability of EDGE capable

terminals. This has already happened as Nokia & Motorola will bring the terminals

by the 2H of 2003. Basically this was obvious after the big US vendors started to

implement GSM-EDGE networks. The only question mark is whether EDGE will be

widely deployed or will it become a niche technology.

This depends much on the early success of UMTS in Europe. If the first

implementations in 1H2003 succeed, which will bring in more network launches in

2H2003 and 1H2004, then EDGE will more probably stay as a niche technology

deployed only to part of the networks. This is because when UMTS succeeds, the

focus of the industry is again shifted to UMTS and the investments on GSM

networks will be minimized. Then EDGE will probably experience similar situation

as HSCSD did. Most probably there will be also operators, which will not go for

UMTS in the near future and for those EDGE offers a feasible solution. Altogether

early UMTS success would make EDGE a niche technology.

If UMTS success won’t happen in short term, due to for example technical

difficulties, then EDGE has a better chance to become a widely deployed

technology. In this case there will be wider range of EDGE terminals available

before UMTS is deployed in larger scale. This will also bring in more EDGE

operators as the investments on network technology are considerably smaller.

The long-term success of EDGE is also dependant on the operators strategy on

driving UMTS with terminal subsidization. This is easily the case in countries where

the license terms are tight and demand a quite rapid and wide UMTS deployment.


49

The most probable case is that EDGE will live alongside UMTS and will in most

cases be used as a geographical complement to UMTS. The UMTS early success is

not very likely, so it will give operators a chance to consider EDGE as well (which is

already ready as a technology) and invest still to their GSM networks. Also as the

technology develops in the future, it will be more likely that EDGE will be included

in the terminals very cheaply in the long run. This makes the multi-mode terminal

manufacturing more feasible to the vendors. So it seems that EDGE will have a

relatively good future ahead, especially in those countries, which don’t have too tight

UMTS license terms for the operators or have granted licenses cheap. In the long run

the standardization must succeed to include the Iu-interface support to GERAN,

otherwise EDGE can easily be positioned as the “poor man’s UMTS” in the market

as similar QoS and services cannot be offered.


50

10. Conclusions

EDGE is relatively easy and cheap to bring to the operators network. It demands for

significantly minor changes to the operational side as well. This means that it can be

quite attractive for an operator, which have not made strong commitments to UMTS.

EDGE uses 8-PSK modulation, which enables approximately 2,5 times the

performance or capacity of GPRS. This new modulation scheme requires new radio

parts for the terminals as well, which means that new terminals must be introduced

from the vendors and the penetration of the EDGE capable terminals grow enough

before this capacity gain compared to GPRS can be fully utilized by the operators.

At the moment there are commitments made from the biggest vendors to introduce

EDGE terminals by 2H2003.

The size of the investments on EDGE depends on the operator. The worst case is that

large part of the network infrastructure is old enough to not support easy EDGE

implementation. The capability of the infrastructure depends of the network vendor.

In some cases the base stations must be fully replaced by newer ones before EDGE

can be implemented. In typical cases only EDGE capable TRXs and BSS software

must be implemented along with changes to Abis interface capacity. If the network

infrastructure is new enough, the base stations can be already equipped with EDGE

capable TRXs. Then only software and enhanced transmission capacity must be

implemented and thus costs can be kept quite minimal.


51

The strategy for an operator to choose on EDGE depends on many things. Most of it

can be even personal (preferences of decision making people) or contractual

(towards vendors) reasons, which drive the network evolution. But in principle one

deciding factor is the commitment towards UMTS and the license terms of the

UMTS license. For those operators without 3G license, EDGE offers a pretty

straightforward business case as a stepping stone before UMTS. For operators with

UMTS license a strategy of EDGE and UMTS co-exist, but for different user

segments, can be useful if EDGE is decided to be deployed as widely as (or wider

than) UMTS. Another strategy can be seen to deploy EDGE as a complementary

solution, to different coverage areas. This is the most feasible strategy if there are

multi-mode terminals available widely.

The future role of EDGE depends very highly on the success of EDGE in USA, the

availability of EDGE terminals, the early success of UMTS, subsidization of UMTS

terminals and the completeness of EDGE standardization.

The most probable case is that EDGE will live alongside UMTS and will in most

cases be used as a geographical complement to UMTS. The UMTS early success is

not very likely, so it will give operators a chance to consider EDGE as well (which is

already ready as a technology) and invest still to their GSM networks. Also as the

technology develops in the future, it will be more likely that EDGE will be included

in the terminals very cheaply in the long run. This makes the multi-mode terminal

manufacturing more feasible to the vendors. So it seems that EDGE will have a

relatively good future ahead, especially in those countries, which don’t have too tight

UMTS license terms for the operators or have granted licenses cheap. In the long run

the standardization must succeed to include the Iu-interface support to GERAN.


References

Auramo 2002 Auramo J. 2002. Enhance Your GSM Networks to 3G with EDGE.

Presentation. Nokia.

Ericsson 2002 http://www.ericsson.com/products/white_papers_pdf/

edge_wp_technical.pdf

Ericsson 2003 www.ericsson.com, accessed 2003

ETSI 2000 ETSI. 2000. ETSI GSM 03.08, V7.4.0, Organization of Subscriber

Data. ETSI.

GSMWorld 2003 www.gsmworld.com, accessed 2003

Halonen et al 2002 Halonen T., Romero J., Melero J. 2002. GSM, GPRS and EDGE

performance - Evolution towards 3G/UMTS. Wiley.

Mouly 1992 Mouly, M. & Pautet. 1992. The GSM System for Mobile

Communications. Published by the authors.

Nokia 2000 Nokia Networks Oy. 2000. NED (Nokia electronic documents) viewer

version 3.11. Nokia Networks Oy.

Nokia 2003 http://www.nokia.com/, accessed 2003

Ovum 2003 http://www.ovum.com/go/ovumcomments/016489.htm, accessed 2003

Penttinen 1999 Penttinen, Jyrki. 1999. GSM-tekniikka; Järjestelmän toiminta, palvelut

ja suunnittelu. (GSM Technique; Function, Services and Planning of

the System, in Finnish). Porvoo: WSOY.

Penttinen 2002 Penttinen Jyrki. 2002.GPRS in Wireless Data. WSOY

Rantanen 2001 Rantanen J. 2001. The Third Generation Cellular Network Solutions

from Operator’s Perspective. Master’s thesis. TKK.

WAPForum 2003 http://www.wapforum.org/, accessed 2003

http://www.wapforum.org/

http://www.ovum.com/go/ovumcomments/016489.htm

http://www.nokia.com/

http://www.gsmworld.com/

http://www.ericsson.com/

http://www.ericsson.com/products/white_papers_pdf/

Positioning_EDGE.doc

Business

edge evolution

gsm network

gprs network architecture

edge radio network planning

gsm system

gprs architecture

edge investment strategies

gsm operatorswith