WCDMA evaluation system—Evaluating the radio … developed the WCDMA evaluation system to...

IntroductionEricsson is fully committed to UMTS/IMT-2000 standardization, the aim of which isto produce a truly global standard for third-generation mobile systems. Ericsson’swideband code-division multiple access

(WCDMA) evaluation system facilitatesUMTS/IMT-2000 standardization, by allowing operators and engineers to demon-strate and test new, advanced, third-generation services and technical solutions.It also enables them to learn about and evaluate WCDMA characteristics in asyn-chronous transfer mode (ATM) environ-ments. Ericsson developed the evaluationsystem• to demonstrate the potential of the third-

generation (WCDMA) system; • to boost intercontinental standardization

processes; • to provide technical input to proposed

new standards; and• to obtain valuable field experience and

input for current commercial develop-ment;

• to pave the way for the evolution of second-generation systems into UMTS.

With the WCDMA evaluation system,Ericsson has taken WCDMA technologyout of the laboratory and put it to test in realradio environments. Thus, Ericsson is nowpoised to exploit and evaluate theWCDMA-related technology of third-generation radio-access systems.

BackgroundLate in the year 1996, Ericsson received aninvitation to tender from NTT DoCoMo—one of the world’s largest mobile phone op-erators—requesting that Ericsson build aWCDMA evaluation system. At the time,however, a standard for WCDMA as a radio-access technology did not exist. In 1997, theJapanese standardization body, the Associ-ation of Radio Industries and Broadcasting(ARIB), began drafting a standard forWCDMA (this first standard was chieflybased on input from NTT DoCoMo in col-laboration with Ericsson).

Ericsson agreed to deliver a test systembased on early proposals to the ARIB stan-dard. The standard has since matured andhas been harmonized with the Europeanstandard from the European Telecommuni-cations Standards Institute (ETSI). The de-tail standardization work continues and is currently being conducted within theframework of the Third-generation Part-nership Project (3GPP).

Work on the system began in early 1997,and although it is sometimes called an “experimental” system, Ericsson’s designersaimed much higher, building what could betermed a pre-commercial system. In fact, the

56 Ericsson Review No. 2, 1999

WCDMA evaluation system—Evaluating the radioaccess technology of third-generation systemsJan Eldståhl and Anders Näsman

The aim of UMTS/IMT-2000 standardization is to produce a truly globalstandard for third-generation mobile systems. Ericsson’s WCDMA evalua-tion system facilitates UMTS/IMT-2000 standardization, by allowing oper-ators and engineers to demonstrate and test third-generation servicesand technical solutions. It also enables them to learn about and evaluateWCDMA characteristics in ATM environments.

Ericsson developed the WCDMA evaluation system to demonstrate thepotential of the third-generation system, to obtain valuable field experi-ence and input for current commercial development, and to pave the wayfor the evolution of second-generation systems into UMTS.

The authors describe Ericsson’s WCDMA evaluation system, the tech-nical details of the system, its background, and the wide range of end-user services it offers.

3GPP Third-generation Partnership ProjectACCH Associated control channelARIB Association of Radio Industries and

BroadcastingATM Asynchronous transfer modeBCCH Broadcast control channelBRI Basic rate interfaceBTS Base transceiver stationCCCH Common control channelDC Direct currentDCCH Dedicated control channelDTCH Dedicated traffic channelDTX Discontinuous transmissionEb/I0 Bit energy per interference power

spectral densityETSI European Telecommunications

Standards InstituteFACH Forward access channelFDD Frequency division duplexGPS Global positioning systemGSM Global system for mobile phone

communicationsIETF Internet Engineering Task ForceIMT-2000 International mobile telecommuni-

cation 2000IPR Internet protocol routerISDN Integrated services digital net-

workITU-T International Telecommunication

Union – Telecommunications Stan-dardization Sector

Iu Denomination used in standardiza-tion work to denote the interfacebetween the BTS and RNC

Iub Denomination used in standardiza-tion work to denote the interface

between the RNC and MSCIWF Interworking functionksps Kilosymbols per secondLAN Local area networkMCPS Megachip per secondMSC Mobile services switching centerMS-SIM Mobile station simulatorPADP Packet service adaptationPBX Private branch exchangePCH Paging channelPRI Primary rate interfacePSTN Public switched telephone networkQPSK Quadrature phase-shift keyingRACH Random access channelRFC Request for commentsRL-ID Radio link IdentityRNC Radio network controllerSDCCH Stand-alone dedicated control

channelSSCOP' Service-specific connection orient-

ed protocol' (A cellular systemadaptation of the SSCOP. Used inthe evaluation system.)

TCH Traffic channelTCP/IP Transmission control protocol/

Internet protocolUADP UDI service adaptationUDI Unrestricted digital informationUER Unit error rateUMTS Universal mobile telecommunica-

tions systemUPCH User packet data channelWBTB Wideband test bedWCDMA Wideband code division multiple

accessWOS WCDMA operations system

BOX A, ABBREVIATIONS

Ericsson Review No. 2, 1999 57

WCDMA evaluation system has been insmall-scale serial production, with close to30 systems produced to date.

System overviewThe WCDMA evaluation system (Figure 1)consists of• the mobile station simulator (MS-SIM)—

note: mobile stations are also sometimesreferred to as mobile terminals;

• the base transceiver station (BTS);• the radio network controller (RNC);• the mobile services switching center

(MSC); and• the WCDMA operations system (WOS).The system was built for experimentingwith the WCDMA radio-access networkand ATM, and for applying these technolo-gies to different wideband services. Conse-quently, the architecture of the MS-SIM,BTS, RNC, MSC and the WCDMA opera-tions system is very general and flexible.

User data and control information be-tween the RNC and BTS are transported viaa preliminary “Iub” interface based on ATMtechnology (“Iub” is the denomination usedin standardization work for the interface be-tween the BTS and the RNC).

Similarly, a preliminary “Iu” interface—

also based on ATM transport technology—is used between the RNC and MSC.

To a great extent, the WCDMA part ofthe evaluation system is based on previouswork conducted at Ericsson research centers,including the wideband test bed (WBTB,1995). The ATM part of the system is builton top of Ericsson’s new ATM based devel-opment platform, Cello, which was de-signed specifically to give good cost/perfor-mance ATM switching and transport in cel-lular systems.1

Obviously, measurement capabilities areessential in an evaluation system. Measure-ment data is thus collected from appropri-ate parts of the system and sent to theWCDMA operations system for logging.The data can then be post-processed for pre-sentation.

Two identical WCDMA evaluation sys-tems have been set up in Stockholm, one ofwhich is mainly used for evaluation pur-poses; the other is used solely for demon-strations.

Although not actually part of theWCDMA system, a global positioning sys-tem (GPS) is used in the evaluation systemto provide different nodes with accuratetime stamps, so that measurement data fromthe nodes can be correlated. The GPS also

BTS

WOS

WOS

WOS

WOSWOS

WOS

MS

ATM router

PBX

IWF

GSM MSC

ISDN converter

ISDN converter

MSCRNC

BTS

BTS

LAN internet/Łintranet

PSTN/ŁISDN

Figure 1System overview of Ericsson’s WCDMAevaluation system.

Microsoft® and NetMeeting® are either regis-tered trademarks or trademarks of MicrosoftCorporation in the United States and/or othercountries.

Intel® and ProShare® are registered trade-marks of Intel Corporation in the United Statesand other contries.

PointCast® is a registered trademark of PointCast Incorporated.

TRADEMARKS

provides positioning data that can be usedfor monitoring the whereabouts of mobilestations.

End-user applicationsThird-generation mobile telephony systemswill offer end-users a wide variety of appli-cations, such as high-speed multimedia ser-vices and other services currently only avail-able from fixed networks. Ericsson’sWCDMA evaluation system enables opera-tors to evaluate and demonstrate third-generation services.

It is difficult to predict what will happennext as the telecommunications, computerand entertainment industries converge. Asmentioned, above, the WCDMA evaluationsystem was developed to meet the demandsof wireless communication in a true multi-media environment. The system has beendeveloped independently of third-generation end-user applications, providinga wide range of bearer services that permitdifferent end-user applications to be evalu-ated (Figure 2).

Voice and audioCertainly, standard voice services currentlyavailable in second-generation systems mustalso function in third-generation systems.The WCDMA evaluation system is thusequipped with 8.8 kbit/s voice codecs

(G.729). It also provides a means of evalu-ating and demonstrating interconnection toexisting GSM networks.

Video conferencingVideo conference or video phone applica-tions based on circuit-switched connections(such as ISDN) or packet-switched media(such as TCP/IP) are being tested—for example, Microsoft Netmeeting or IntelProshare-based products. Other video ap-plications like remote monitoring are alsopossible.

Still- and moving-picture codecs arebeing tested on the WCDMA evaluationsystem.

Internet applicationsThe WCDMA evaluation system supportsbest-effort packet-switched data up to ap-proximately 470 kbit/s. Access to the Internet and corporate intranets is essentialto professional mobile users.

“Push applications” (such as PointCast)are becoming increasingly popular on theInternet, allowing users to subscribe to spe-cific kinds of information; for example,weather forecasts, advertisements, news,company bulletins, and updates of compa-ny data, such as price lists. These applica-tions usually entail a background type oftraffic; that is, the information reaches theuser even when he or she is occupied withother applications. Packet-switched servicesin the evaluation system support back-ground traffic. A mobile terminal can belogged onto a server and only pay for up-dated information sent out and not for theconnection time.

Corporate LAN accessToday, many people remotely access fileservers, databases, and groupware applica-tions on their company LAN. For good per-formance, high-speed connections must beoffered that are at least as fast as landlinemodems operating at 56 kbit/s.

When working remotely, corporate userswant to be “on-line” with their company sys-tem; for instance, to see when they receivee-mail. Since cellular calls currently costmore than landline calls, subscribers cannotafford to remain connected eight hours aday. Instead, they must dial in several timesa day to check their mail. With WCDMApacket-switched services, however, it will bepossible to charge subscribers for volume ofdata transferred instead of for the durationof the session.


Internet applications • WWW browsing • Video telephone • E-mail • News push • Networked games

Voice • 8 kbit/s • Call between WCDMA terminal and GSM core network

Internet/ intranet

WCDMA network

Application server

Video terminal

Wireless postcard and business card • Tailor-made demo application

GSM MSC

Video conference • High-quality tailor-made applications • Standard ISDN package

PSTN/ŁISDN

Figure 2End-user applications in the evaluation system.

Partners cooperating with Ericsson to testthird-generation wireless capabilities

Market PartnerCanada Microcell (GSM Alliance)China Third-generation R&D cooper-

ation with China Academy ofTelecommunications Technol-ogy

Germany Mannesmann Mobilfunk and T-Mobil

Italy Telecom Italia MobileJapan Japan Telecom

NTT DoCoMoUK Testbed open to all UK-based

operatorsHong Kong SmarToneSweden Telia

TABLE 1, COOPERATING PARTNERS


Wireless postcard and electronicbusiness cardsE-mail has become one of the most commonmobile data applications on cellular networks.E-mail often includes more than plain-textmessages, however. Many messages also carryattachments of additional files—for instance,files of images and video clips. E-mail of thiskind requires greater transmission speedsthan the present 9.6 kbit/s.

Electronic business cards are currentlybeing standardized—in addition to name,title, company address, and so on, they mayinclude images and other information.Thus, in the future, instead of seeing thenumber of a caller, his or her photographmight be displayed on the WCDMA mo-bile terminal (which also doubles as a digi-tal camera).

Other applicationsDifferent user groups have different needsas relates to mobile multimedia. For exam-ple, the health sector is investigating inter-active health and medical applications(telemedicine); the security sector is inter-ested in remote monitoring; and the trafficsector works with traffic telematics and nav-igation equipment in vehicles. Further-more, fire brigades and broadcasting com-panies could benefit from WCDMA systemsto assess situations as the first fireman orjournalist comes on the scene.

Intelligent living involves video-on-demand, on-line entertainment applica-tions, and applications that support work-ing and shopping from home. These areaswill also require mobile applications.

Worldwide deploymentFollowing NTT DoCoMo’s request for aWCDMA evaluation system, Ericsson received several other queries from operatorsaround the world (Table 1). Ericsson wanted to make the evaluation system avail-able to as many operators as possible. In theUnited Kingdom, for example, Ericsson operates a test system that is open to all op-erators—today, every major UK-based operator is working with Ericsson and sev-eral new operators have also shown interest.

System architecture

Base transceiver stationTo handle mixed services, the base trans-ceiver station in third-generation systems

must have a flexible architecture. To meetthe requirements for flexibility, designersstructured the hardware architecture according to function. That is, instead ofusing a channel-based architecture, as wasdone in first- and second-generation systems(Figure 3),the BTS in Ericsson’s WCDMAevaluation system uses a pooled architec-ture. This solution provides greater flexi-bility for coping with the varied demandsof future services (Figure 4).

COMBBB

BB

BB

TRXs ANT

BB

RF

RF

RF

RF

PA

PA

PA

PA

Figure 3Architecture of first- and second-generation base transceiver stations.BB BasebandPA Power amplifierRF Radio frequency

MCPABB

BB

BB

ANT

BB

RF

RF

RF

RF

Figure 4Architecture of the third-generation base transceiver station.MCPA Multiple carrier power amplifier

The antennas, which are passive units fortransmission and reception, do not requirea direct current (DC) power supply. Twoidentical antennas are used in each sector:one is used for both transmission and re-ception; the other is solely used for recep-tion. Instead of two antennas in each sector,a dual-polarized antenna can be used, there-by reducing the number of antennas to oneper sector. The antenna for each macrocellsector has a mechanical tilt and a fixed elec-trical tilt.

The architecture of the base station per-mits the bearer capacity to be freely allocat-ed to users of different types of service: • voice service;• circuit-data service carried as unrestrict-

ed digital information (UDI) up to384 kbit/s;

• raw packet-data service up to 472 kbit/s(packet throughput capacity is dependenton radio interference, and consequently,on the SSCOP' retransmission rate).

Limiting factors of the total user data band-width are the data capacity of the mobilestation, the radio cell size, and radio inter-ference.

Radio network controllerThe RNC (which in GSM terminology cor-responds to the BSC) is built on Ericsson’sgeneric ATM infrastructure. The processorand all other devices are connected to theATM switch, which makes it very easy toextend system capacity. The RNC houses• the radio resource and macro diversity-

combination function (also referred to asthe soft handover combination function);

• the speech coding/decoding function according to ITU-T recommendationG.729 (provided the codecs are located inthe RNC);

• the “Iub” transport based on 1.5 or2 Mbit/s ATM links with ATM adapta-tion layer 2 (AAL2) link termination;

• the “Iu” interface (carried on a 155 Mbit/sATM link); and

• timing and synchronization functions.At least three base stations can be connect-ed to one RNC, each with up to two 1.5 or2 Mbit/s “Iub” links. The traffic capacity ofthe example in Figure 6 is approximately160 mobile stations connected to each otheror to the fixed network. Every connection isswitched in the MSC.


PWC

MIO

MP

TU

ATM

MMXALT

ATM-IFXET 1.5

To WOS Ethernet

To WOS Ethernet

-48VDC

ATM-IFX

ATM-IFC

ATM-IFC

TRX- DIG

TRX- RF

MCPA

LNA

To RNC 1.5/2 Mbit/s

TX/RXANT

RXANT

ENC

DEC BBRX BBIF

BBRA

BBTX

switch

ATM Baseband TRX AMP

Figure 5Architecture of the BTS.ALT ATM link termination BBRX Baseband receiverATM-IFC ATM interface client BBTX Baseband transmitter LNA Low-noise amplifierATM-IFX ATM interface host DEC Decoder MMX ATM multiplexorBBIF Baseband interface ENC Encoder TRX-DIG Transceiver, digital partBBRA Baseband random access ET Exchange terminal TRX-RF Transceiver, radio frequency part


The architecture of the RNC is very flex-ible, making it ideal for experiments withadvanced WCDMA radio-network func-tions. For instance, to meet new or increaseddemands, additional printed circuit boardsmay be installed in empty board positionsin RNC subracks. The RNC can also handle large amounts of measurement datain real time, which capability is necessaryfor experimenting with advanced radio-network functions.

Mobile services switching centerLike the RNC, the MSC is built on Ericsson’s generic ATM switch infrastruc-ture and has the same flexible properties asthe RNC. Indeed, certain functionality canbe moved between the RNC and MSC. Thecodecs may be put in either the MSC or theRNC—for example, to comply with “Iu” in-terface standardization, or for other experi-mental purposes.

The MSC handles the “Iu” interface to theRNC and the fixed network interfaces toISDN and ATM LAN. The main task of theMSC is to set up calls to and release callsfrom mobile stations. It includes only a rudi-mentary level of signaling sequence over theair interface for location updates, authenti-cation, ciphering, and so on. Because it canhandle voice as well as circuit-switched and

PWC

MIO

MP

TU

RCU

ATM

DHTALT

ET 155ET 1.5

To WOS Ethernet

To WOS Ethernet

-48VDC

Clock ref

External network

CODEC

To BTS 1.5/2 Mbit/s

”lu” 155 Mbit/s ATM

switch

RNC

Figure 6Architecture of the RNC.PWC Power connectionRCU Reference clock unitMIO Multi-purpose input/outputMP Main processorTU Timing unitDHT Diversity handover

PWC

MIO

MP

ET 155

ET PRI ISDN converter

ATM LAN

EC IWF To GSM/MSC A i/f (MAP/ISUP)

BRI (2B+D)

PRI (30B+D)PRI (23B+D)

155 Mb POTS

TU

ATM

UADPPADP

IPR

ET 155

To WOS Ethernet

To WOS Ethernet

-48VDC

To RNC 155 Mbit/s

switch

MSC

Figure 7Architecture of the MSC.EC Echo canceller

packet-switched data communication with-in the same node, the MSC facilitates experimentation with interesting multime-dia applications. The MSC includes • the echo canceller function;• the IP router;• the data adapter functions for circuit- and

packet-data services;• ISDN and adapter function interface;• ATM LAN interface; and• the speech coding/decoding function

according to ITU-T recommendationG.729 (provided the codecs are located inthe MSC).

External equipment, such as an ISDN con-verter, a private branch exchange (PBX) andan ATM router, is used to adapt user data

to analog voice, circuit-data and packet-datanetworks.

The external primary rate interface(ET PRI) complies with the Japanese stan-dards, JT-Q.931 (Layer 3), JT-Q.921 (Layer2) and JT-I.431a (Layer 1). Moreover, whenan external ISDN converter is connected, theEuropean ISDN interfaces (PRI and BRI)can also be provided according to the ITU-T recommendations, Q.931 (Layer 3),Q.921 (Layer 2) and I.430/I.431 (Layer 1).The external ATM LAN uses a permanentvirtual circuit (PVC) to carry all IP-over-ATM connections according to theIETF RFC 1483 specification.

Interworking functionality (the IWFunit) in the MSC facilitates connection to aGSM/MSC, which by inserting the GSMsubscriber’s SIM card into the MS-SIM(SIM-card roaming), enables the MS-SIM toreceive voice calls made to a GSM sub-scriber. The IWF provides a simplified in-terface according to the GSM A-interface.

The built-in Internet protocol router hasthe capacity to route IP packets at rates upto 10 Mbit/s. Distribution of capacity be-tween voice, circuit data and packet data canbe provided by different configurations ofthe echo cancellor, UDI service adaptation(UADP), packet service adaptation (PADP),Internet protocol router (IPR) and the PRI.

Mobile station simulatorThe MS-SIM was designed exclusively foruse in the evaluation system and differs


PWC

IPR

PADP

OSC

ATM

MMX

MP

TU

UADP

ATM-IFXCODEC

ETPRI

ATM-IFX

ATM-IFC

ATM-IFC

Analog speech

UDI 1.5 Mbit/s

Packet data Ethernet

-48VDC

TX/RX ANT

RX ANT

ENC BBRX BBRA

BBIFBBTX

DEC

switch

Controller Baseband

TRX-DIG

TRX

WOS Ethernet

Duplex Duplex

TRX-RF

Figure 8Architecture of the mobile station simula-tor (MS-SIM).

MS-to-MS calls: • 80 voice connections;• 24 UDI connections at 64 kbit/s;• 3 UDI connections at 384 kbit/s;• 16 packet connections at 76 kbit/s;• 3 packet connections at 472 kbit/s; or • a mix of these services.MS-to-external calls: • 16 voice connections and 24 UDI connctions at 64kbit/s; or• 6 UDI connections at 384 kbit/s over the primary rate interfaces.MS-to-packet-data services:• 32 ATM LAN IP connections at 76 kbit/s; or• 6 connections at 472 kbit/s.

TABLE 2 TRAFFIC CAPACITY OF A TRIPLE BASE STATION SYSTEM


considerably from a commercial unit. It wasbuilt drawing on solutions and componentsfor the BTS and design solutions for theRNC/MSC (Figure 8).

ATM CelloThe BTS, RNC, MSC and MS-SIM havebeen built on the ATM switch platform,which handles internal communication(within each node) as well as communica-tion between nodes and all external connec-tions.1

System characteristicsThe system operates within the IMT-2000 band and has the following charac-teristics:• Operational band—uplink (reverse)

1,920-1,940 MHz; downlink (forward)2,110-2,130 MHz;

• 190 MHz duplex distance;• Frequency division duplex (FDD);• 4,096 Megachips per second (Mcps) chip

rate; • 5 MHz carrier width;• Modulation/demodulation method

– for data: QPSK, pilot-aided coherentdetection RAKE receiver;

– for spreading: QPSK;• Encoding/decoding method

– innercoding for traffic channels andACCH;

– convolutional encoding (R=1/3,K=9), soft-decision Viterbi decoding;

– innercoding for all control channels ex-cept ACCH;

– convolutional encoding (R=1/2,K=9), soft decision Viterbi decoding;

– bit interleaving (all channels);• Short code, 256 to 32 chip long-layered

orthogonal code; and• Long code

– downlink, 10 ms (use 216 –1 chip longGold code cut into 10 ms lengths);

– uplink, 225 · 10 ms (use 241 –1 chip longGold code cut into 225 · 10 ms lengths).

Every channel in the WCDMA network canbe concentrated to a single sector. Con-versely, channels can be freely distributedover three sectors and two carriers.

All ATM transport functionality—inter-nally in the BTS and for the “Iub”—is de-rived from the Ericsson ATM switch infra-structure in the RNC and MSC.

The MSC can handle 16 codecs for a totalof 16 fixed-network voice connections.

In each BTS, it is possible to connect traf-fic capacity corresponding to

• 6 sectors with one 5 MHz carrier; or • 3 sectors with two 5 MHz carriers per sec-

tor. For circuit-switched transmission it is pos-sible to connect • 64 voice connections; • 16 UDI connections at 64 kbit/s; or• 2 UDI connections at 384 kbit/s.For packet-switched transmission, it is pos-sible to connect • 8 packet-switched connections at

76 kbit/s; or• 2 packet-switched connections at

470 kbit/s. It is also possible to mix services; for example:• 16 voice channels;• 4 UDI connections at 64 kbit/s;• 2 packet-switched data connections at

76 kbit/s; and• 1 packet-data connection at 472 kbit/s.

The air interfaceThe air interface is based on early versionsof what is today being finalized in the 3GPP.The carrier width is 5 MHz. Thus, given thetotal bandwidth of 20 MHz, there are fourselectable frequencies. Each carrier containsseveral physical channels (Figure 9), each ofwhich is defined by a unique• long code;

Physical channels

Dedicated

Common

Common control

256 ksps

Perch 16 ksps

32 ksps

128 ksps

16 ksps

64 ksps

Figure 9 Physical channels.

• short code; and• transmission rate.A variety of information is transmitted be-tween the BTS and the mobile station. Dif-ferent logical channels are used depending onthe type of information being transmitted.Each logical channel is used for a specific pur-pose, such as paging, call setup and speechtransmission. The logical channels are mappedonto the physical channels (Figure 10).

Logical channelsThe air interface consists of three kinds oflogical channel, namely:• common control channels (CCCH);• dedicated control channels (DCCH); and• traffic channels (TCH).

Common control channels

The broadcast control channel (BCCH 1/2) isa unidirectional downlink channel for trans-mitting system control information on cellsor sectors—such as long code phase infor-mation, uplink interference power, and soon—from the BTS to the mobile station.This channel transmits system informationwhose content changes over time. The in-formation is transmitted continuously.

The paging channel (PCH) is a unidirec-tional downlink channel that transfers pag-ing information from the BTS to the mobilestation. The same paging information canbe transmitted simultaneously via severalBTSs as well as to several sectors of each BTS.

Random access channels (RACH-short,RACH-long) are unidirectional uplinkchannels for receiving initial access requestsfrom mobile stations.

Forward link access channels (FACH-short,FACH-long) are unidirectional downlinkchannels for transmitting control informa-tion and random access acknowledgmentsfrom the BTS to a mobile station. The com-mon RACH/FACH channels are also usedfor transmitting low-speed packet data toand from several mobile stations.

Dedicated control channels

The stand-alone dedicated control channel(SDCCH) is a point-to-point bidirectionalchannel that transfers control informationbetween the mobile station and the RNC.The channel occupies one physical channel.The SDCCH exchanges information on ser-vice selection, authentication, and locationupdates. The SDCCH is only used after ran-


Perch channel (16 ksps)

Physical channels Logical channels

Physical channel for common control (64 ksps)

Physical channel for common control (16 ksps)

Dedicated physical channel (32 ksps)



BCCH1

UPCH

ACCH

DTCH

SDCCH

RACH-S

RACH-L

FACH-L

FACH-S

PCH

BCCH2

Figure 10Mapping of logical channels onto physicalchannels.


dom access. After use, the connection is re-leased.

The associated control channel (ACCH) is apoint-to-point bidirectional channel thattransfers control information between a mo-bile station and the RNC. This channel isassociated with the dedicated traffic chan-nel (with which it is multiplexed on a single physical channel).

Traffic channels

The dedicated traffic channel (DTCH) is apoint-to-point bidirectional channel thattransmits and receives user information.

Discontinuous 32 ksps transmission issupported for speech. To support high-speed packet-switched data, the multicodefunction associates one radio-link identi-ty (RL-ID) with multiple codes of identi-cal symbol rates. (See also Table 3) This isthe case when a high-speed service (384 kbit/s) is transmitted on three 256 ksps physical channels—each with aunique code.

The user packet data channel (UPCH) is apoint-to-point bidirectional channel be-tween the MSC and a mobile station that isused for packet-data services (user packet

Logical channels

Traffic TCH

Control CCH

Common ctrl CCCH

Dedicated ctrl DCCH

SDCCH

ACCH

UPCH-C

DTCH

UPCH-U

FACH-S/L

RACH-S/L

BCCH1

BCCH2

PCH

Figure 11Structure of the logical channels.

Service Information rate Logical channel Physical channel rateVoice 8 kbit/s DTCH 32 kspsUDI 64 64 kbit/s DTCH 128 kspsUDI 384 384 kbit/s DTCH 3·256 kspsPacket 76 76 kbit/s UPCH 128 kspsPacket 470 470 kbit/s UPCH 3·256 ksps

TABLE 3 CORRELATION BETWEEN SERVICES, LOGICAL CHANNELS AND PHYSICAL CHANNEL RATE

data and control packet data). The multi-code function associates one RL-ID withmultiple codes of identical symbol rates(valid for 3 · 256 ksps).

EvaluationNumerous tests can be executed within theEricsson WCDMA evaluation system. Thedescriptions that follow highlight the capa-bilities of the system.

Transmitter and receiver performanceA large battery of tests is used for evaluat-ing transmitter and receiver performance.The tests cover basic performance under“normal” conditions, quantitative testsmade in the laboratory using specified chan-nels, and field trials, which are assumed tobe slightly more demonstrative of “real-life”characteristics. Some tests may also be per-formed under extraordinary and rapidlychanging conditions (for instance, when auser enters or exits a building). These testsevaluate the robustness of algorithms andthe effects of delay in the system.

Detector performanceDetector performance is tested to verifysearcher/tracker performance in differentenvironments, to find suitable parametersettings for observed and anticipated multipath environments, and to assess per-formance during soft handover, where syn-chronous performance can be closely moni-tored. The detector performance tests• verify equipment sensitivity;• investigate searcher performance in dif-

ferent dispersive environments;• test tracker performance for known dis-

persion types;• verify searcher/tracker performance dur-

ing soft/softer handover;• measure Eb/I0 detectors for different

interference types; and• investigate the effects of using different

numbers of rake fingers.

Idle modeTests of the idle mode indicate the generalcharacteristic performance of specific algo-rithms, radio interface and mobile-stationhardware implemented in the evaluationsystem. The tests also assess parameter de-pendency and the performance of sector se-lection, including time delay while in idlemode. In particular, the tests examine ini-tial and continuous sector selection in idlemode.

Open-loop performanceMeasurements can be made to estimate theextent to which measured data on one linkcan be used for predicting the accuracy ofanother link when setting various parame-ters. The measurements, which can be madein stationary and variable environments, es-timate the quality of the open-loop powercontrol for the RACH/FACH and the ini-tial power setting when setting up the CCH.The tests of open loop performance predict:• power level capability; and• the consequences of RACH/FACH and

the initial CCH power setting.

Random access performancecharacteristicsCertain tests evaluate the performance of therandom-access procedure implemented inthe WCDMA evaluation system as well asthe effect of different parameter settings ofthe acquisition procedure. The tests measurethe effects of interference levels and the in-terfering effects that RACH/FACH proce-dures have on established calls. In particu-lar, the tests• assess RACH performance in different

power-regulation schemes; and• examine disturbances from

RACH/FACH bursts.

Uplink and downlink controlTesters evaluate the performance of the fastuplink/downlink power control loop interms of speed, stability, performance whenEb/I0 is low, and operation during soft/soft-er handover.

The practical performance of the controlloop, in terms of its ability to maintain aconstant Eb/I0 value is in variable radio environments, is assessed for estimating theneed for reference value control. A specialarea of interest focuses on performance at cellfringes and, in this context, the effects ofsoft/softer handover. The effects of suddenenvironmental changes (turns in the road,outdoor/indoor usage) can also be studied,as well as step responses during handoverand resynchronization. The tests also checkthe effects of low data rates; for instance, dueto poor synchronization or when DTX isused. The downlink tests are similar to testsof the uplink power control. The tests of theuplink and downlink power control assess• response times;• the effects of simultaneous multiple chan-

nel types;• the influence of different dispersion envi-

ronments;



• the effects of changing environment;• the effects of TPC channel quality;• performance during soft/softer handoff;• the influence of DTX; and• the effects of mobile terminal velocity.

Outer loopThe outer loop or quality control loop is usedto compensate for changes in statisticalproperties as well as the effects of imperfectdiscriminators and processing. The controlloop can be verified for different multipath(dispersion) environments and changes be-tween such environments. The tests mayalso be used for studying how well the outerloop compensates for non-ideal behavior ofthe fast loops in the uplink and the down-link. The tests of the outer loop assess• the dynamic range/granularity required

for a specific Eb/Io target;• single channel performance;• the effects of changing environment;• performance during soft/softer handover;

and• the effects of mobile station velocity.

Soft/softer handoverTesters evaluate the performance of differ-ent soft/softer handover criteria in real radioenvironments, including quasi-stationaryand rapidly changing radio environments(for example, at street corners and streetcrossings. They also assess performance as afunction of different parameter settings(timers, weights), investigate perch powerlevels and estimate soft handover gain. Thetests of soft/softer handover• identify critical parameters for sector de-

tection and selection timing;• assess the effects of different evaluation

criteria;• assess the influence of active set parame-

ters;• assess the influence of perch power levels

and other parameters;• assess handover gain;• assess performance during cell breathing;

and• assess synchronization.

Hard handoverTesters evaluate the performance of the sys-tem’s hard handover function at differentparameter settings. In particular, they as-sess the effects of sector selection and syn-chronization. The tests of hard handover as-sess• the importance of different parameters;• synchronization;

• hard handover gain; and• the effects of cell breathing.

Radio switching (packet handover)Testers evaluate the performance of the sys-tem’s radio-switching function, particular-ly during “normal” behavior. The tests ofradio switching assess • the effects of different evaluation criteria;• reporting intensity;• quality and parameter sensitivity;• radio switching setup times;• street effects; and• synchronization.

Channel type switchingTesters evaluate the performance of channeltype switching when different evaluationcriteria and reporting intensities are used.They also study handover quality, which in-cludes the effects of synchronization and thehandover procedure. The effects of cellbreathing may also be studied, provided asuitable test environment can be generated.Observed traffic behavior can be comparedwith that of traffic models. The tests of chan-nel type switching assess• switching intensity/parameter sensitivity;• quality;• street effects;• the relevance of available traffic models;• differences between coverage of dedicat-

ed and common channels; and• the effects of cell breathing.

Call setup performanceCall setup performance is studied for differ-ent services and terminals. Testers studyvoice, UDI 64 kbit/s and 384 kbit/s as wellas packet-data traffic between mobile sta-tions (MS-to-MS) and between mobile sta-tions and the PSTN, ISDN and GSM net-works. The tests of call setup performanceassess• the performance of different services; and• the influence of mobile stations.

ATM transport characteristics Testers evaluate various aspects and charac-teristics of ATM transport performance. Ifthe capacity of the transport network far ex-ceeds that of the air interface, system per-formance is not significantly affected. How-ever, if the capacity of the transport networkis on a par with that of the air interface, itwill have a noticeable impact on system per-formance. The main focus of the test casesis on performance, measured as delay andframe loss versus throughput. The tests also

examine functionality and synchronization.The tests of ATM transport characteristicsassess• the dependence of load on throughput and

quality; • synchronization; and• statistical multiplexing of packet data and

voice, which is used to reduce the demandfor transport capacity.

Protocol measurementsTests of protocol measurements study pro-tocol performance, measured as unit errorrate (UER), peer-to-peer delay and rate ofretransmission of single protocols, and in-teractions between different protocol lay-ers. These interactions are particularly in-teresting as they relate to the study ofpacket applications that retransmit onmany protocol layers. The issues of timingand synchronization, as they relate to therespective system functions, are covered bytests of those functions. The purpose ofthese test cases is to study the delay, tim-ing, and synchronization properties of thesystem that are not directly associatedwith any particular system function. The

protocol measurements help testers to as-sess• the performance of a single protocol; and• interaction between different protocols.

End-user application characteristicsTests of end-user application characteristicsare made to study perceived end-user quality. These tests focus on physical measurements, UER, round-trip delay, andso on.

Issues that interest testers include data rateefficiency. For example, are high data-trans-mission rates significantly better than lowerdata-transmission rates when channel typeswitching and its consequences on retrans-mission on different layers is taken into ac-count relative to the actual duration of the call?Studies of end-user application characteristics • compare the efficiency of different data

rates; and• assess service quality, including the Web.

Radio network characteristics/radionetwork planningTests of radio-network characteristics andplanning help testers to estimate system be-


Superframe (640 ms)

Radio frame (10 ms) Radio frame #0

Time slot (0.625 ms)

Pilot symbols Symbols for logical channel

Long code mask symbol

TPC symbol

Perch channel 16 ksps

Downlink physical channel for common control 64 ksps

Dedicated physical channel 256 ksps



Time slot #1

4 5 1

4 5

4 36

8 1

1

1

151

4 75

4 15

1

Time slot #2 Time slot #16

1st perch channel Perch 1

2nd perch channel Perch 2

Radio frame #1 Radio frame #63


Figure 12Physical channel format—every physicalchannel has a three-layer structure thatconsists of superframes, radio frames andtime slots. The structure of the radioframes and time slots varies, dependingon the physical channel and symbol rateused.


havior, including coverage/capacity in dif-ferent configurations. The tests • identify stability limits;• assess near/far effects;• assess perch power level planning;• verify orthogonality with mixed channels;• verify path-loss prediction (half-screen);

and• assess the effects of pulsed interference.

Antenna solutionsTesters evaluate the influence that differentantenna types and tilting have on coverageand soft handover. The tests of antenna so-lutions assess• the effects of practical antennas; and• the usefulness of antenna tilting.

ConclusionEricsson began testing WCDMA during theearly part of 1995. Today, Ericsson operatesfully developed WCDMA radio networks inseveral parts of the world.

The architecture of the MS-SIM, BTS, RNC,MSC and the WCDMA operations system isstructured by function and is very flexible.

All ATM transport functionality is de-rived from the Ericsson ATM switch infra-structure in the RNC and MSC.

The WCDMA evaluation system wasbuilt for experimenting with theWCDMA radio-access network and ATM,and for applying these technologies to different wideband services. Thus, opera-tors can use the system to evaluate nu-merous WCDMA system characteristicsincluding• transmitter and receiver performance;• detector performance;• idle mode performance;• open loop performance;• random access performance;• uplink power control;• downlink power control;• soft/softer handover;• hard handover;• radio switching (packet handover);• channel type switching;• call setup performance;• ATM transport;• protocol measurements;• end-user applications; and • antenna solutions.

Radio frame (10 ms)

Time slot (0.625 ms)


Uplink physical channel for common control (RACH-L) 64 ksps

Time slot #1

4 36

Time slot #2 Time slot #16

Radio frame (10 ms)

Time slot (1.25 ms)

Pilot symbols (4 symbols)

Pilot symbols (4 symbols)


Uplink physical channel for common control (RACH-S) 16 ksps

Time slot #1

4 16

Time slot #2 Time slot #8 Figure 13The format of the physical channel forcommon control differs slightly from thatshown in Figure 12, since it handlesaccess (random access and packet dataon common channels) for several mobilestations.

1 Reinius, J.: Cello—An ATM transport andcontrol platform. Ericsson Review Vol.76(1999):2, pp. 48–55.

REFERENCES

WCDMA evaluation system—Evaluating the radio … developed the WCDMA evaluation system to...

Documents

Transcript of WCDMA evaluation system—Evaluating the radio … developed the WCDMA evaluation system to...