Tu XiaodongLi Yazhuo Taofeng
(W ireless Technology Innovation Institute (WTI),Beijing University of Posts and Telecommunications, Beijing 100876, China)
Abstra c t:The FuTURE 4G Tim e Division Dup lex(TDD)trial system uses 3.5 GHz carrier frequency and several crucial technolog ies inc lud ing b roadband Multip le Input Multip le Output(MIMO)and Orthogonal Frequency Division Multip lexing(OFDM).These technologies challenge the link budget and networking analysis of the FuTURE 4G TDD trial network.This paper analyzes the p ractical 3.5 GHz p ropagation m odel and the link budget of Rad io Frequency(RF)param eters of the trial system.Moreover,it introduces networking analysis and network p lanning of the trial system,which com bines the field test results of the MIMO system.The FuTURE 4G TDD tria l system and its trial network have been accomp lished w ith successfu l checkup.The trial system fulfills a ll the requirements w ith two Access Points(AP)and three Mobile Term ina ls(MT),which supports mu lti-user,m obility,a high peak rate of 100 Mb/s,High-Definition TV(HDTV),high-speed data download,and Voice over IP(Vo IP)services.
This w o rk was funded by the NationalNatural Science Foundation of China under Grant 60496312,and by the 863 Prog ram of China under Gran ts 2003AA12331004 and 2006AA01Z260.
The research and developmentof nextgenerationmobile communication systems has d rawnmuch attention worldw ide.Among other issues,researchers'main concern is how tofurther imp rove the spectrum efficiency and offer subsc ribers w ith betterexperience ofservices.
China has begun its research on next generationmobile communication system s since 2001 and has developed Future Technology forUniversalRad io Environment(FuTURE)p rojec t,which is a partof the NationalHigh Technology Research and DevelopmentProgram of China(i.e.,863 Prog ram)[1].This p roject is designed to develop a testing p latform for4Gmobile communication systems and its trialsystem s in China,w ith the aim s ofstudying the trends and demands in the field ofw ireless communication in the next10 years.The p rojec talso has the goals ofdevelop ing technologies fornextgeneration communication system s,and estab lishing related c ritical technology verification system so as to support future-oriented,new w ireless communication services and to p lay an important role in 4G standard ization.
Two branches are involved in the FuTURE p lan:4G Frequency Division Dup lex(FDD)system and 4G Time Division Dup lex(TDD)system[2].
The FuTURE 4G TDD trialsystem,consisting of two Access Points(AP)and three Mobile Term inals(MT),was accom p lished in June 2006.Itp rovides a multi-usernetworking testenvironment thatsupportsmobility,and adop ts severaladvanced technologies inc lud ing Multip le InputMultip le Output(MIMO),OrthogonalFrequency Division Multip lexing(OFDM),Soft Frac tional Frequency Reuse(SFFR)and generalized d istributed network architecture[3].
Accord ing to the networking requirements of the FuTURE 4G TDD trial system,it is required tomake the link budgetand conductnetworking analysis tomake betternetwork p lanning after the jointdebugging of the trialsystem and before the construction and networking testof the trialnetwork.
The FuTURE 4G TDD Trialsystem uses 3.5 GHz carrier frequency w ith central frequency of3.45 GHz,but the study on the p ropagation characteristics of3.5 GHz rad io waves has juststepped into the startup period in China,and in other countries as well,and no general p ropagationmodels have been developed.
Besides,the MIMO technology used in the trialsystem has high requirements forantenna installation location and radio p ropagation environment.
Therefore,the link budgetand networking analysis of the trialsystem are critical to the construction of the trial network.
▼Table 1. RF parameters of FuTURE 4G TDD trial system
This paperd iscusses the link budget and the field test-based networking analysis of the trialsystem in detail.
Beforemaking the link budgetof the FuTURE 4G TDD trialsystem,itis necessary to determ ine the related parameters ofRad io Frequency(RF)equipmentof the system,especially the transm itpower,receiver sensitivity,noise figure,antenna pattern,antenna height,and gain.In add ition,as MIMO technology is used,the configuration and geometric d istribution ofantennas should also be taken into account.
The detailed RFparameters of the FuTURE 4G TDD trialsystem are listed in Tab le 1.
In order to dep loy the link budgetof the FuTURE 4G TDD trialsystem,we should firstcom pute the coverage area based on the p ropagationmodel,which is set up according to the p ropagation
characteristics of3.5 GHz rad io waves.Then the coverage of the FuTURE 4G TDD trialsystem can be ob tained.
3.5 GHz Radio Waves
The FuTURE 4G TDD trialsystem uses 3.5GHz carrier frequency,w ith central working frequency of3.45GHz and 20MHz bandw id th.Currently,the common models used to study the p ropagation characteristics of rad io waves are Okumura and Hata.And the Hata modelis furtherd ivided into two kinds:the Okumura-Hatamodel for frequencies from 150MHz to 1 500MHz and the Cost231-Hata,an extension ofOkumura-Hata,for frequencies up to 2GHz[4-5].However,due to the com p lexity of the ac tual rad io environment,no theoreticalmodelcan perfectly suita specific situation,and nomodelcan be app lied to allp ropagation environments.As a result,in actual app lication,a p ropermodelhas to be selec ted fora specific environment.
As the FuTURE 4G TDD trialsystem uses 3.45GHz carrier frequency,neither Okumura norHatamodel,which covers frequencies nomore than 2 GHz,is app licab le here.Nowadays,the study of the p ropagation characteristics of3.5GHz radio waves has attracted more and more attention in China,as wellas in other countries.The reason for this comes from the c rowded situation at the carrier frequency of2GHz or lower,and the futuremobile communication system smay have towork athigher frequencies.The research on the p ropagation characteristics of3.5GHz and 5GHz rad io waves has been launched recently.
In recentyears,the Wireless Technology Innovation Institute(WTI)of Beijing University ofPosts and Telecommunications(BUPT)has donemuch research on the p ropagation characteristics of3.5GHz radio waves;ithas also conducted many field tests and measurements in the urban environmentaround BUPT.Afteranalyzing a large numberof field testdata,the WTIhas setup a 3.5GHz-based p ropagationmodel[6-7].
▲Figure 1. The BUPT-centered field test area.
Figure 1 shows the BUPT-centered testarea forstudying the p ropagation characteristics of3.5GHz rad io waves.The b lack point in the center is the antenna installation location,which is on the top of themain building ofBUPT.Three antennas are configured in the test:one forsend ing and twofor receiving.The rad ius of the testarea exceeds 2 km,and moving speed is up to 100 km/h.
Tab le 2 lists the average path loss fac tors of the fourd irections.These values are allcalculated from the field testdata.The testcenter is at the antenna installation location.
The follow ing link budgetof the FuTURE 4G TDD trialsystem is based on the abovemodel.As the trialnetwork is centered on the cam pus ofBUPT,the model from the BUPT-centered field test data is no doubtan accurate and p roperone.
Based on themodel,the com puting formula for the path loss of3.5 GHz band can be exp ressed as follows:
▼Table 2. Path loss factors
Where PL(d)is the path loss through a d istance of d(unit:km);λ is the waveleng th;and d0is a reference d istance,forwhich 10m is used.To make the link budgetmore reliab le,a certainmargin should be reserved.So,in actualcalculation,the largestpath loss factor n(i.e.,North;see Tab le 2)is used.
Another im portantparameter involved in the link budget is receiver sensitivity,which is the lower lim it for the receiver to correctly receive and identify signals.To ensure good communication,the transm itted signal,which w illexperience space attenuation when traveling,must be correctly identified at the receiver.Thatis to say,the signalstrength at the receivermustbe g reater than the receiver sensitivity.
In the FuTURE 4G TDD trialsystem,themaximum transm itpowerofup link MT is smaller than thatofAP,making the trialsystem up link lim ited.Therefore,the coverage of the system should be calculated based on the up link.
The receiver sensitivity atAP end is calculated as follows:
Accord ing to the performance index requirements,to ensure a BitError Rate(BER)of10-6,the required Eb/N0should be 3 dB.When the hardware overhead is taken into consideration,the required Eb/N0shallbe converted into 12.3 dB of Signal-to-Noise ratio(SNR)(the trial system adop ts 16QAM mode and 1/3 cod ing).Thatmeans the outputSNR(SNRout)is 12.3 dB.From Tab le 1,the noise figure(F)of the AP receiver is 5 dB.So the required inputSNR(SNRin)at the AP receiverend should be:SNRin=SNRout+F=17.3 dB (2)
The thermalnoise power can be com puted using the formula:N =K TB,where K is the Boltzmann constantwhose value is 1.381×10-23W/Hz/K,T is the room temperature(290 K),and B is the bandw id th(20MHz).The noise power is:
From Tab le 1,we can see the antenna gain atAPend(GAP)is 14 dBi.Therefore,them inimum signalpower thatcan be received by the AP(i.e.,the receiver sensitivity)is:Prmin=SNRout+F-GAP+N=-98.9 dBm(4)
Once the receiver sensitivity and the maximum transm itpowerofMTare known,themaximum path loss allowed by the system can be calculated out.
Accord ing to Tab le 1,the antenna gain atMT(GMT)is 5 dBi,and the maximum transm itpoweratMTend is 27 dBm when the BER is required to be 10-6.So,the theoreticalmaximum path loss allowed by the system is:
Where Lotheris the loss ofother devices in the trialsystem,such as feeder,which is about2 dB.In p ractice,shadow fading p rotection margin and fast fading p rotectionmargin should be taken into consideration.By experience from actualengineering[8-9],the estimated fast fad ing p rotection margin is 3 dB,and the shadow fad ing p rotec tionmargin is 7 dB.Therefore,the maximum path loss allowed by the system is:
Rep lace the PL(d)in equation(1)w ith the resultofequation(6),and equation(7)is ob tained:
So themaximum up link coverage radius of the FuTURE 4G TDD trialsystem can be calculated:dmax=0.303 km (8)Asmentioned above,this trial system is up link lim ited.To ensure service continuity and smooth handoverofan MT from one AP to another,the coverage of the trialsystem should be no more than themaximum up link coverage radius.
With the above link budget calculations,we can go on to the networking analysis of the FuTURE 4G TDD trialnetwork.The trialnetwork is builtw ith its center located at the BUPT campus,and configured w ith two APs and three MTs to support multi-user,mobility and a peak data rate of100Mb/s.
Although the p ractical 3.5GHz-based p ropagationmodel from the study of the p ropagation characteristics in BUPT is used in the link budgetcalculation,the theoreticalresult may stilld iffer from the ac tualradio environment,especially in a case where MIMO technology is used,which is sub jec t to antenna layout,installation and rad io environment.Therefore,in add ition to link budget,the networking analysis of the actualrad io environmentis necessary to betterp lan the two APs'installation locations,the spacing between them,and the handoverarea.
The basic RF parameters used in the real rad io environment field test for the FuTURE 4G TDD trialsystem are listed in Tab le 3.For the purpose ofhardware debugging,a sim p lified configuration of two transm itand four receiverantennas is app lied in the MIMO system for the field test.The two transm itantennas are 120°d irectional,w ith amaximum transm itpowerof27 dBm;while the four receiverantennas are omni.At the transm itterend,the antennas are laid in a horizontalarray,w ith a spacing of1.5m in between and a tiltang le less than 15 deg rees,so thata highergain can be obtained.At the receiverend,the antennas are also laid in a horizontal array,butw ith a spacing of1m.
▼Table 3. Basic RF parameters for the field test of radio environment
▲Figure 2. Field test Result of the radio environment in BUPT.
In the field test,the receiver judges the communication quality based on the Block ErrorRates(BLER)of the data it receives.The test is conduc ted on the BUPT campus,w ithin a rectangulararea of350m×200m.This area is a built-up area surrounded w ith buildings and trees thatactas reflectors,d iffrac tors or scatters ofsignals during the test.
After severalmeasurements w ith the testvehic lemoving in the area,the average data in Figure 2 are ob tained.
As the field testuses actualMIMO links tomeasure the rad io environment,the measurement results are valuab le materials forstudying the coverage of the MIMO system,as wellas the layoutof the antennas.
In general,the MIMO system perform s quite wellin the test.At the transm itpowerof27 dBm,the coverage is about200m,and the BLER can reach its bestvalue of0.Inmostpartof the area,the signal recep tion is quite good,w ith the BLER less than 5%(thatis to say,BER is between 10-5and 10-6).The trees and build ings in the area have little impacton the system.Even in the case ofscattering and diffraction,the system works well,w ith a BLER of less than 5%.
A FuTURE 4G TDD trial network on the BUPT cam pus is builtbased on the above link budgetand testdata analysis.
The network,consisting ofone Control Unit(CU),two APs and three MTs,supports multi-userand mobility management like handover.The coverage radius ofeach AP reaches 200m.With a BER of10-6,a BLER of less than 5%and a peak data rate of 122Mb/s,the system can supportHigh-Definition TV(HDTV),high-speed data download,and Voice over IP(Vo IP)services.
The FuTURE 4G TDD trialsystem uses 3.5 GHz carrier frequency and several key technologies inc lud ing MIMO and OFDM.Configured w ith two APs and three MTs,the system supports multi-user,mobility and a peak data rate of100Mb/s.Atp resent,the trialsystem has been estab lished and app roved.The link budgetand networking analysis have p layed an important role in the successfulconstruction of the trial network.Moreover,the data ob tained from the link budgetand the field testare usefulreferences for the network construction ofnextgenerationmobile communication system.