《通信电子线路》课程教学资源（教材文献）Circuits and Systems for Wireless Communications（PDF电了书，Edited by Markus Helfenstein and George S. Moschytz）.pdf_P16-P20

CIRCUITSANDSYSTEMSFORWIRELESSCOMMUNICATIONS1PartIhasbroughttogetherkeyengineersfrom industryand academiatoshed lightonperformancedemands,boardlevel design,andsub-micron silicon CMOS solutionstakingthereaderthroughrealisticdesignscenariosforRFsystemintegration.Devicesandsystems,suchaspagers,cellularand cordlessphones,cablemodemsmobile faxes, PDAs (Personal Digital Appliances), wireless LANs, and RF identifica-tion tags are rapidly penetrating all aspects of our lives, evolving from luxury itemsto indispensabletools.Semiconductor and system companies,small and large,ana-logueanddigitalhaveseenthestatisticsandarestrivingtocapturetheirownmarketshareby introducing various RF products.Today's pocket phones contain more thanonemillion transistors,with only a small fraction operating in theRF range and therest performing low-frequency“base-bandanalogue and digital signal processingHowever, the RF section is still thedesign bottleneck ofthe entire system.In contrast to other types of analogue and mixed-signal circuits, RF systems de-mandnotonlyagoodunderstandingofintegrated circuits,butalsoofmanyareasthatare not directly related to integrated circuitsCommunicationTheoryMicrowaveRandomTheorySignalsSignalRFTransceiverPropagationDESIGNArchitecture1MultipleIC DesignAccessWirelessCADStandardsToolsFigure1.1 RFdesigndisciplines.Most of the areas shown in Fig.1.1 have been studied extensively for more thantwo decades,making it difficult for an IC designer to acquire the necessaryknowledgein a reasonable amount oftime.Traditional wireless systemdesign has thus been car-ried out at somewhatdisjointed levelsofabstraction:communication theorists createthemodulation schemeandbase-bandsignalprocessing,RFsystemexpertsplanthetransceiver architecture:IC designers developeach ofthebuildingblocks:andmanufacturersglue"theICsandotherexternalcomponentstogether.Infact,architecturesare oftenplannedaccordingtothe availableoff-the-shelfcomponents,and ICsarede-signedto serve as many architectures as possible,leading toa great deal ofredundancyat both system and circuit levels.This results in higher levels of power consumptionandgenerallylowerperformance

4 CIRCUITS AND SYSTEMS FOR WIRELESS COMMUNICATIONS Part I has brought together key engineers from industry and academia to shed light on performance demands, board level design, and sub-micron silicon CMOS solutions, taking the reader through realistic design scenarios for RF system integration. Devices and systems, such as pagers, cellular and cordless phones, cable modems, mobile faxes, PDAs (Personal Digital Appliances), wireless LANs, and RF identification tags are rapidly penetrating all aspects of our lives, evolving from luxury items to indispensable tools. Semiconductor and system companies, small and large, analogue and digital, have seen the statistics and are striving to capture their own market share by introducing various RF products. Today’s pocket phones contain more than one million transistors, with only a small fraction operating in the RF range and the rest performing low-frequency “base-band” analogue and digital signal processing. However, the RF section is still the design bottleneck of the entire system. In contrast to other types of analogue and mixed-signal circuits, RF systems demand not only a good understanding of integrated circuits, but also of many areas that are not directly related to integrated circuits. Most of the areas shown in Fig. 1.1 have been studied extensively for more than two decades, making it difficult for an IC designer to acquire the necessary knowledge in a reasonable amount of time. Traditional wireless system design has thus been carried out at somewhat disjointed levels of abstraction: communication theorists create the modulation scheme and base-band signal processing; RF system experts plan the transceiver architecture; IC designers develop each of the building blocks; and manufacturers “glue” the ICs and other external components together. In fact, architectures are often planned according to the available off-the-shelf components, and ICs are designed to serve as many architectures as possible, leading to a great deal of redundancy at both system and circuit levels. This results in higher levels of power consumption and generally lower performance

RFSYSTEM INTEGRATION5Most recently,as theindustrymoves towardhigher integration and lower costRFand wireless design increasingly demands more“concurrent engineering."thereby re-quiringICdesignersfrombothindustry andacademiatocombineforces andtohaveasufficient and integrated knowledge of all the disciplines [1]RF circuits must processanaloguesignalswithawidedvnamicrangeathighfreguencies.Itisinterestingtonote that the signals must be treated as analogue even if the modulation is digital ortheamplitudecarriesnoinformation.COMPLEXITYSUBMICRONTECHNOLOGYNOISEPOWERCONSUMPTION工FREQUENCYSPEEDSUPPLYVOLTAGEDISTORTIONXVOLTAGEGAINSWINGFigure1.2RFdesignhexagon.Thetrade-offsinvolved inthedesignof suchcircuitscanbe summarised inthe“REdesign hexagon"shown in Fig.12.While any of the seven parameters tradeto someextent,all these parameters are severelyconstrained bythecoreparameters,namelythepowerconsumptionand supplyvoltage.Itisimportanttorecognisethat.whiledigitalcircuitsdirectlybenefitfromadvancesinIC technologies.RFcircuitsdonotbenefit as much.This issue is exacerbated by the fact that RF circuits often requireexternalcomponents-forexample.inductors-thataredifficulttobringontothechipeven in modernIC processes.RFdesign techniques arethus becoming highly sensi-tiveto device physics,and so analog characterisation of digital VLSI technology isof primary concern.Oneof themajorchallenges is implementing RFcircuits oniCsinsteadofPCBs,offering advantages including lower production cost, highfunctionality,small physical size,highreliability,andlow power requirements.Itnowbecomesvery necessary to achieve better co-ordination between the“system design"activityand the“"RF circuitdesign"activity.A few years ago Gallium-Arsenide (GaAs) technology was the primary-choicesemiconductorfor implementingRFICs dueto its lownoisefigure,highergain andhigher output power.Advances in sub-micron silicon CMOS, however, have madeit possible to achieve higher levels of RF system integration at lower cost than withGaAs, predominantly for low-GHz-band wireless applications [2].The other benefitsofCMOSRF arethegreatermanufacturability andminimised powerrequirements to

RF SYSTEM INTEGRATION 5 Most recently, as the industry moves toward higher integration and lower cost, RF and wireless design increasingly demands more “concurrent engineering,” thereby requiring IC designers from both industry and academia to combine forces and to have a sufficient and integrated knowledge of all the disciplines [1]. RF circuits must process analogue signals with a wide dynamic range at high frequencies. It is interesting to note that the signals must be treated as analogue even if the modulation is digital or the amplitude carries no information. The trade-offs involved in the design of such circuits can be summarised in the “RF design hexagon” shown in Fig. 1.2. While any of the seven parameters trade to some extent, all these parameters are severely constrained by the core parameters, namely the power consumption and supply voltage. It is important to recognise that, while digital circuits directly benefit from advances in IC technologies, RF circuits do not benefit as much. This issue is exacerbated by the fact that RF circuits often require external components—for example, inductors—that are difficult to bring onto the chip even in modern IC processes. RF design techniques are thus becoming highly sensitive to device physics, and so analog characterisation of digital VLSI technology is of primary concern. One of the major challenges is implementing RF circuits on ICs instead of PCBs, offering advantages including lower production cost, high functionality, small physical size, high reliability, and low power requirements. It now becomes very necessary to achieve better co-ordination between the “system design” activity and the “RF circuit design” activity. A few years ago Gallium-Arsenide (GaAs) technology was the primary-choice semiconductor for implementing RF ICs due to its low noise figure, higher gain and higher output power. Advances in sub-micron silicon CMOS, however, have made it possible to achieve higher levels of RF system integration at lower cost than with GaAs, predominantly for low-GHz-band wireless applications [2]. The other benefits of CMOS RF are the greater manufacturability and minimised power requirements to

6CIRCUITSANDSYSTEMSFORWIRELESSCOMMUNICATIONSdrive off-chip loads. While integrated silicon BJT transceivers are still more desirablefor today's products, CMOS RF solutions are looking very promising,with the real-istic prospect of a single-chip transceiver in a plastic package.Furthermore, newerdevicetechnologies suchasSiliconGermanium arematuringrapidly andofferthehigh mobility necessary for today's RF wireless products.This array of competingtechnologies offers systemdesignersmore creativeopportunity,and the best wirelesstransceiver solutions may well emerge from system design evolving together with ar-chitecture,circuits,antennas, and power allocation plans.In thefuture,base-bandsignal processing will inevitably make up for imperfections in the front end (e.g. soft-wareradio),PartI begins with a section by Gordon Aspin from TTP Communications, a com-panywithvastexperienceinRFsystemintegrationforcellularproducts.Thesectiondescribes the realistic design ofa part of a fully integrated transceiver IC from Hitachiwhich satisfiesmulti-band GSMRF specifications.Some of the subtletiesofboard-level integration are presented, coupled with a design approach which attempts tomake practical GSMhandset design a more straightforward task.Emphasis is placedupon the importance ofunderstanding total system-level requirements when designinga chip, and upon how board level design influences low-level requirements.In the next section,Peter MolefromNortel Semiconductors gives an overviewofsystem integration on a chip.Peter takes us through a number of practical RF design is-sues and thendiscussesgeneral problems that radio systems must overcome toachieveacceptableperformance.The section overviewspractical concerns forbothreceiverand transmitter and howdifferent radioarchitectures can beutilised toovercome someoftheproblems.Thesectionconcludeswithanumberofpracticaldesignissuesforintegratingradiocircuitryin silicontechnology.Thefinal two sections concentrateon thedesign offully integratedtransceiverchipsin sub-micron and deep sub-micron Silicon CMOS technologies. Michiel Steyaertfrom theKatholiekeUniversiteit of Leuven introduces us to the arena ofusing deepsub-micron CMOS to create single-chiptransceiver blocks and components such asLNAs,VCOs,up-converters,synthesisers etc.to satisfy cellular performance spec-ificationsabove1GHz.ThesectiondiscussesallthebottlenecksandchallengesofRFCMOSusingplaindeepsub-microndevicesforintegration within systems suchasDECT,GSM,andDCS1800.Finally,QiutingHuang et.al.fromtheIntegrated Systems Laboratoryat ETHZurich presents apractical high-performanceGSM transceiverfront-end in a 0.25μmCMOS process.This section concludes PartIby taking the reader through a practicalRFsystem integrationexample.TheworkdemonstratesthatexcellentRFperformanceisfeasiblewith0.25um CMOS,even intermsoftherequirements ofthesuper-heterodyne architecture.Designforlownoise and lowpowerforGSMhandsets has beengivenparticularattentionIn conclusion,PartI will give the reader a practical evaluation of state-of-the-artRFsystemdesign and integrationfor GHz wirelesscommunications.Thechapters inPartiencompass the failures, successes, and most ofall the realistic RF challenges toenabletotal integration of portablefuture wireless information systems

6 CIRCUITS AND SYSTEMS FOR WIRELESS COMMUNICATIONS drive off-chip loads. While integrated silicon BJT transceivers are still more desirable for today’s products, CMOS RF solutions are looking very promising, with the realistic prospect of a single-chip transceiver in a plastic package. Furthermore, newer device technologies such as Silicon Germanium are maturing rapidly and offer the high mobility necessary for today’s RF wireless products. This array of competing technologies offers system designers more creative opportunity, and the best wireless transceiver solutions may well emerge from system design evolving together with architecture, circuits, antennas, and power allocation plans. In the future, base-band signal processing will inevitably make up for imperfections in the front end (e.g. software radio). Part I begins with a section by Gordon Aspin from TTP Communications, a company with vast experience in RF system integration for cellular products. The section describes the realistic design of a part of a fully integrated transceiver IC from Hitachi which satisfies multi-band GSM RF specifications. Some of the subtleties of boardlevel integration are presented, coupled with a design approach which attempts to make practical GSM handset design a more straightforward task. Emphasis is placed upon the importance of understanding total system-level requirements when designing a chip, and upon how board level design influences low-level requirements. In the next section, Peter Mole from Nortel Semiconductors gives an overview of system integration on a chip. Peter takes us through a number of practical RF design issues and then discusses general problems that radio systems must overcome to achieve acceptable performance. The section overviews practical concerns for both receiver and transmitter and how different radio architectures can be utilised to overcome some of the problems. The section concludes with a number of practical design issues for integrating radio circuitry in silicon technology. The final two sections concentrate on the design of fully integrated transceiver chips in sub-micron and deep sub-micron Silicon CMOS technologies. Michiel Steyaert from the Katholieke Universiteit of Leuven introduces us to the arena of using deep sub-micron CMOS to create single-chip transceiver blocks and components such as LNAs, VCOs , up-converters , synthesisers etc. to satisfy cellular performance specifications above 1 GHz. The section discusses all the bottlenecks and challenges of RF CMOS using plain deep sub-micron devices for integration within systems such as DECT, GSM, and DCS 1800. Finally, Qiuting Huang et. al. from the Integrated Systems Laboratory at ETH Zurich presents a practical high-performance GSM transceiver front-end in a 0.25 µm CMOS process. This section concludes Part I by taking the reader through a practical RF system integration example. The work demonstrates that excellent RF performance is feasible with 0.25 µm CMOS, even in terms of the requirements of the super-heterodyne architecture. Design for low noise and low power for GSM handsets has been given particular attention. In conclusion, Part I will give the reader a practical evaluation of state- of-the-art RF system design and integration for GHz wireless communications. The chapters in Part I encompass the failures, successes, and most of all the realistic RF challenges to enable total integration of portable future wireless information systems