A Simulation Tool for Third Generation CDMA Systems Presentation to IEEE Sarnoff Symposium

Transcription

1 A Simulation Tool for Third Generation CDMA Systems Presentation to IEEE Sarnoff Symposium March 22, 2000 Fakhrul Alam, William Tranter, Brian Woerner Mobile and Portable Radio Research Group () URL: This work was carried out with sponsorship of industrial affiliates and LGIC

2 Outline of Presentation Introduction : Evolution of Third Generation (3G) Systems Description of WCDMA Physical Layer Description of the Simulator Simulation Results Performance Enhancing Features Conclusion & Extension of Research 2

3 Introduction The goal for Third Generation Systems: Communication to anybody, anywhere, anytime. Intended Services High Speed Data Video and Multimedia Traffic Voice Signals 3G communicators are oriented towards multimedia message capability marking a significant leap from current standards. 3

4 Earlier Generations of Cellular Systems First Generation Analog Frequency Modulation (FM) Frequency Division Multiple Access (FDMA) Voice Traffic Advanced Mobile Phone System (AMPS): most popular 1 st Generation system Second Generation Digital Counterpart of First Generation Digital Modulation Schemes with compression and coding Time Division Multiple Access (TDMA) and Code Division Multiple Access (CDMA) along with FDMA Voice Traffic USDC standards IS-54 & IS-136, GSM, PDC, cdmaone 4

5 Third Generation Systems Mobile phone meets Internet A picture tells a thousand words : but it needs a lot of bandwidth 5

6 Primary Requirements of Third Generation Systems Voice quality comparable to Public Switched Telephone Network (PSTN) Support of high data rate Vehicular : 144 kbps Pedestrian : 384 kbps Indoor Office : 2 Mbps Backward compatibility with pre-existing networks and flexible introduction of new services and technology More efficient usage of the available radio spectrum Adaptive radio interface suited to the highly asymmetric nature of most Internet communications 6

7 Evolution of 3 G PDC ARIB (WCDMA) TDD GPRS WCDMA Multicarrier Multicode GSM UTRA (WCDMA) FDD EDGE 136 HS AMPS IS-54 ANSI UWC-136 cdmaone cdma2000 IS-95B This Figure is condensed from Figure 4, page 28 of Malcom W. Oliphant, " The Mobile Phone Meets the Internet," IEEE Spectrum, pp , August

8 WCDMA Key nical Characteristics Multiple Access Scheme Duplex Scheme Multirate/Variable rate scheme Chip Rate Carrier Spacing Frame Length Inter Base Station synchronization DS-CDMA FDD/TDD Variable spreading factor and multi-code 3.84 Mcps MHz (200 khz carrier raster) 10 ms FDD: No accurate synchronization needed TDD: Synchronization required Channel Coding Scheme Convolutional Code (rate 1/2 and 1/3, const. length 9) Turbo code 8

9 WCDMA Physical Channel Structure Dedicated Channels : Connection dedicated Dedicated Physical Data Channel (DPDCH) Dedicated Physical Control Channel (DPCCH) Common Channels : Shared among users Primary and Secondary Common Control Physical Channel (CCPCH) at the Downlink Synchronization Channel (SCH) at the Downlink. 9

10 Uplink Frame Structure (DPDCH & DPCCH) Pilot Data TFCI FBI TPC T slot =2560 chips, 10 2 k bits (k = 0..6) Pilot for Coherent Demodulation and Channel Estimation TFCI (Transport Format Combination Indicator) to Indicate and Identify Several Simultaneous Services FBI (Feedback Information) to support techniques requiring feedback Slot 1 Slot i Slot 15 T f =10 ms TPC (Transmit Power Control) for power control purposes SF 256 = k 2 Frame 1 Frame i Frame 72 T super =720 ms 10

11 Downlink Frame Structure (DPDCH / DPCCH) TFCI Data 1 TPC Data 2 Pilot T slot =2560 chips, 10 2 k bits (k = 0..7) Slot 1 Slot i Slot 15 T f =10 ms SF k = Frame 1 Frame i Frame 72 T super =720 ms 11

12 Uplink Spreading & Modulation (DPDCH & DPCCH) Channelization Code(C D ) cos(ω c t) Data (DPDCH) Scrambling Code(C SC ) p(t) Control (DPCCH) j p(t) Channelization Code(C C ) sin(ω c t) Mobile Station Specific (MS) Complex Scrambling Code MS has one DPDCH and one DPCCH 12

13 Downlink Spreading & Modulation (DPDCH / DPCCH) cos(ω c t) Scrambling Code(C SC ) p(t) DPDCH/ DPCCH Serial to Parallel Channelization Code(C ch ) j p(t) sin(ω c t) Base Station (BS) Specific Complex Scrambling Code Each user has one DPDCH and one DPCCH 13

14 OVSF Codes Orthogonal Variable Spreading Factor (OVSF) codes preserve orthogonality between channels Each level in the code tree defines channelization codes of length SF All codes of the same level constitute a set : orthogonal to each other Any two codes of different levels are orthogonal if one code is not the mother of the other c 1 (1)= (1) c 2 (1)=(1,1) c 2 (2)= (1,-1) c 4 (1)= (1,1,1,1) c 4 (2)= (1,1,-1,-1) c 4 (3)= (1,-1,1,-1) c 4 (4)= (1,-1,-1,1) SF = 1 SF = 2 SF = 4 14

15 OVSF Codes (cont d) All codes within the code tree cannot be simultaneously used within one BS/by one MS Perfect orthogonality is maintained only at zero lag Auto-correlation peak is not narrow : difficulty in synchronization Auto-correlation of the 3 rd code (SF=256) 15

16 Scrambling Codes Maintain separation among different MS/BS Two types Short : Advanced receivers employing Multi User Detection (MUD) or Interference Cancellation Long:Employed in the Simulator C 1 C 2 2 C 2 j C SC C C ( w jw C ) sc = (w 1 is a repetition of {1,1, } at the chip rate ) w 2 w 2 = { } Csc = C + jw C C C ( 2k) = C ( 2k + 1) = C ( 2k)

17 Scrambling Codes (Cont d) Uplink Generator Polynomials: Downlink Generator Polynomials: X X X X X X Generate sequences x and y Generate Gold Code sequences C 1 and C 2 from x and y 1 Cross-correlation Distribution 17

18 Summary of WCDMA Modulation Spreading Modulation Data Modulation Spreading Scrambling Frame Length Chip Rate Pulse Shaping Dual Channel QPSK for UL Balanced QPSK for DL BPSK for UL QPSK for DL OVSF codes spreading factor for UL spreading factor for DL Complex Scrambling 10 ms 3.84 Mcps Raised Cosine with 0.22 roll off 18

19 WCDMA Simulator Implements the physical layer channels of the WCDMA system Multipath time varying channel Rake diversity combining at the receiver Structured Multiple Access Interference (MAI) AWGN at the receiver front end Study of BER performance under different practical channel condition is possible 19

20 WCDMA Simulator(Cont d) Flexible: Can be very easily modified to include error correction coding antenna diversity Modular: Can be modified to incorporate different receiver schemes different channels 20

21 Uplink Simulator 1 st Interferer Raised Cosine Filter Time Varying Channel 2 nd Interferer Raised Cosine Filter Time Varying Channel N th Interferer Raised Cosine Filter Time Varying Channel + MAI Desired MS Raised Cosine Filter Time Varying Channel + Rake Receiver BER Counter Gaussian Noise Generator Root-Raised Cosine Filter 21

22 Downlink Simulator Frame generator of the 1 st interferer Frame generator of the 2 nd interferer Frame generator of the N th interferer + MAI Frame generator of the desired user + Raised Cosine Filter Time Varying Channel + Rake Receiver BER Counter Gaussian Noise Generator Root-Raised Cosine Filter 22

23 Channel Parameters (Proposed by ETSI) Relative Dealay (ns) Avg. Power (db) Relative Dealay (ns) Avg. Power (db) Indoor Channel Indoor to Outdoor Channel Relative Dealay (ns) Avg. Power (db) Vehicular A Outdoor Channel 23

24 Time Varying Channel Rayleigh Waveform Rayleigh Waveform Delay of the 2nd path Transmitted Frame Rayleigh Waveform Delay of the Nth path 24

25 Implementation of Delay N-τ samples τ samples N samples = 1 Frame τ samples N-τ samples N samples = 1 Frame 25

26 Conversion of Tabulated Channel Parameters with Ray Splitting Splitting of ray 1 Power Splitting of ray 2 Sample Addition of the two rays 26

27 Frequency Response of the Indoor Channel 27

28 Frequency Response of the Outdoor Channel 28

29 Addition of Noise and Pulse Shaping Square Root Raised Cosine Filter Channel Square Root Raised Cosine Filter Gaussian Noise Generator Square Root Raised Cosine Filter Square Root Raised Cosine Filter Channel Gaussian Noise Generator Square Root Raised Cosine Filter Raised Cosine Filter Channel Gaussian Noise Generator Square Root Raised Cosine Filter 29

30 Rake Receiver e -jϕ1 Frame Alignment Resampling Descrambling Despread z 1 e -jϕ2 Frame Alignment Resampling Descrambling Despread z 2 Received Frame MRC e -jϕn Frame Alignment Resampling Descrambling Despread z N 30

31 Graphic User Interface (GUI) GUI Implemented in MATLAB Type wcdma at the MATLAB prompt Choose Between Uplink and Downlink from the Main Menu 31

32 Uplink Menu Run the simulator with default parameters E b /N 0 : 5 db Spreading Factor : 32 Samples per Chip :1 Pulse Shape : RC Channel : Indoor Interferer Number : 0 Rake Finger : 4 Number of Frame : 1 Define Parameters 32

33 Final Uplink Menu Display BER plot Display BER as a text file Plot the multipath profile Plot the data 33

34 Uplink BER vs E b /N 0 at Indoor Channel (User SF=32) BER interferer 2 interferers 4 interferers 8 interferers 12 interferers E b /N 0 in db 34

35 Uplink BER vs E b /N 0 at Outdoor Channel (User SF=32) BER interferer 2 interferers 4 interferers 8 interferers 12 interferers E b /N 0 in db 35

36 Uplink BER vs Interferers at Indoor Channel (E b /N 0 =12 db) BER sf=32 sf=16 sf=8 sf= Number of Interferers 36

37 Uplink BER vs Interferers at Outdoor Channel (E b /N 0 =12 db) 10 0 sf=32 sf=16 sf=8 sf= BER Number of Interferers 37

38 Downlink BER vs E b /N 0 at Indoor Channel (User SF=32) BER interferer 2 interferers 4 interferers 8 interferers 12 interferers E b /N 0 in db 38

39 Downlink BER vs E b /N 0 at Outdoor Channel (User SF=32) interferer 2 interferers 4 interferers 8 interferers 12 interferers 10-1 BER E b /N 0 in db 39

40 Downlink BER vs Interferers at Indoor Channel (E b /N 0 =12 db) BER sf=32 sf=16 sf=8 sf= Number of Interferers 40

41 Downlink BER vs Interferers at Outdoor Channel (E b /N 0 =12 db) 10-1 BER 10-2 sf=32 sf=16 sf=8 sf= Number of Interferers 41

42 Summary of Results The system is interference limited at both the links for higher number of users As the system load approaches 50%, the performance for the uncoded system becomes unacceptable Orthogonality among channels is preserved better at the downlink 42

43 Implementation of Convolutional Coding at the Uplink Voice application at the Uplink Data rate 8 kbps : 80 bits per 10ms frame Add 16 bits of CRC : 96 bits per 10ms Resultant data rate : 9.6 kbps Apply rate 1/3, constraint length 9 convolutional coding. 9-1 = 8 tail bits : = 104 bits per 10ms = 312 bits per 10ms Nearest matching : 300 bits per 10ms Puncture 12 bits 300 bits per 10ms corresponds to a spreading factor of

44 BER/FER vs E b /N 0 for the 9.6 kbps Uplink Service with Coding Indoor Channel 10 0 BER FER 10-1 BER/FER E b /N 0 in db 44

45 BER/FER vs E b /N 0 for the 9.6 kbps Uplink Service with Coding Outdoor Channel 10 0 BER FER 10-1 BER/FER E b /N 0 45

46 Performance Improvement with Coding Target BER : 10-3 (No MAI ) Channel Required SNR for Uncoded System Required SNR for Coded System Indoor 6.8 db 2.1 db Vehicular A 11.5 db 6.75 Coding gain of approximately 4.7 db 46

47 Performance Enhancing Features Adaptive Antennas (at the Receiver) Connection dedicated pilot bits as training sequence Increased capacity and coverage Transmit Diversity (Downlink) Transfer processing burden to the base station Open Loop Time Switched Transmit Diversity (TSTD) Space-Time Transmit Diversity (STTD) Closed Loop Selection Transmit Diversity (STD) Feedback Mode Transmit Diversity 47

48 Performance Enhancing Features (Cont d) Advanced Receivers Multi User Detection and Interference Cancellation Use Short Scrambling Codes for lower complexity Space-Time Rake or 2 D Rake : Combine spatial diversity with temporal diversity 48

49 Conclusion The research has produced a study of the physical layer of WCDMA BER performance has been investigated for different channel conditions Channel coding was implemented The developed simulator can be a useful tool to evaluate performance of WCDMA systems Paper and Presentation will be available at: 49

50 Extension of Work The Simulator is being used to investigate Transmit Diversity Schemes Performance of Adaptive Antenna Array at the Receiver Space-Time Rake Receivers 50