Hệ thống lai ghép CDMA đa sóng mang-đa mã Luận văn ThS. Kỹ thuật điện tử

TR I H C QU C GIA HÀ N I NG I H C CÔNG NGH HOÀNG V N TU N H TH NG LAI GHÉP CDMA A SÓNG MANG- A MÃ Ngành: Công ngh i n T - Vi n thông Chuyên ngành: K thu t i n t Mã s : 60.52.70 LU N V N TH C S NG IH NG D N KHOA H C: PGS.TS NGUY N VI T KệNH Hà n i 2008 M cl c Danh m c ký hi u ch vi t t t Danh m c hình v b ng L i nói đ u Ph n I: Gi i thi u H th ng CDMA lai ghép 11 1.1 K thu t tr i ph CDMA 11 1.2 H th ng CDMA đa sóng mang 13 1.2.1 Mô t h th ng truy n d n a sóng mang 13 1.2.2 Mô hình h th ng lai ghép CDMA đa sóng mang 17 1.2.3 u nh c m c a h th ng MC CDMA 22 1.2.3.a u m c a MC CDMA 22 1.2.3.b Nh 23 c m c a MC CDMA 1.3 H th ng CDMA đa mư 24 1.3.1 Gi i thi u t ng quan 24 1.3.2 Các lo i mã s d ng h th ng CDMA 24 1.3.2.a Mã Walsh 24 1.3.2.b Mã Gold Kasami 25 1.3.3 Mô hình h th ng CDMA đa mã Ph n II: H th ng lai ghép CDMA đa sóng mang đa mã 27 30 2.1 Mô hình h th ng MTC-MC CDMA 30 2.2 Phân tích h th ng MTC-MC CDMA 33 2.3 Hi u n ng c a h th ng MTC-MC CDMA 38 2.4 S thích nghi t c đ c a h th ng MTC-MC CDMA kênh fading đa đ ng 42 2.4.1 Mô hình kênh fading đa đ 2.4.1.a Kênh fading đa đ 2.4.1.b a đ ng 42 ng n hình ng tán x đa đ 2.4.1.c Mô hình hóa kênh fading đa đ 42 ng r i r c 47 ng 47 2.4.2 S thích nghi t c đ 52 2.4.3 Mô ph ng 55 K t lu n 62 Tài li u tham kh o 63 Ph l c: Mư ngu n m t vài ch ng trình mô ph ng (Matlab) 65 Danh m c ký hi u ch vi t t t 3G Third Generation Thông tin di đ ng th h 4G Fouth Generation Thông tin di đ ng th h AWGN Additive White Gaussian Noise T p âm Gauss tr ng c ng tính BCK Biothogonal Code Keying Khóa mư l BER Bit Erorr Rate T l l i bit BS Base Station Tr m g c CDM Code Division Multiplexing Ghép kênh phân chia theo mã CDMA Code Division Multiple Access a truy nh p phân chia theo mư Cdma2000 Code Division Multiple Access 2000 (American 3G Standard) a truy nh p phân chia theo mư theo chu n 3G c a M CP Cyclic Prefix Ti n t vòng DFT Discrete Fourier Transform Bi n đ i Fourier r i r c DS Direct Sequence Chu i tr c ti p DS-CDMA Direct Sequence CDMA CDMA chu i tr c ti p EGC Equal Gain Combining T h p đ l i cân b ng FDM Frequency Division Multiplexing Ghép kênh phân chia t n s FDMA Frequency Division Multiple Access FFT Fast Fourier Transform Bi n đ i Fourier nhanh FH Frequency Hopping Nh y t n IC Interference Cancellation Tri t nhi u ICI Inter-Carrier Interference Nhi u xuyên sóng mang IDFT Inverse Discrete Fourier Trasform Bi n đ i ng c Fourier r i r c IFFT Inverse Fast Fourier Transform Bi n đ i ng c Fourier nhanh IID Independent Identically Distributed Phân b đ c l p đ ng nh t IS-95 Interim Standard-95 Chu n t m th i n m 95 c a CDMA ISI Inter-Symbol Interference Nhi u xuyên ký hi u LTV Linear Time - Varying Tuy n tính thay đ i theo th i gian LTI Linear Time - Invariant Tuy n tính b t bi n theo th i gian MAI Multiple Access Interference Nhi u đa truy nh p ng tr c giao a truy nh p phân chia t n s MBCK Multi-Code Biorthogonal Keying Khóa mư l MC-CDMA Multicarrier CDMA CDMA đa sóng mang MC-DS-CDMA Multicarrier-DS-CDMA DS-CDMA đa sóng mang MCM Multi-Carrier Modulation i u ch đa sóng mang MC-SS Multi-Carrier Spread Spectrum Tr i ph đa sóng mang MF Matched Filter B l c hòa h p M-FSK M-ary Frequency Shift Keying Khóa d ch t n M-m c MIMO Multiple Input and Multiple Output Thi t b nhi u đ u vào – ML Maximum Likelihood Lu t h p lý c c đ i MLSE Maximum Likelihood Sequence Estimation MLSSE Maximum Likelihood Symbol-bySymbol Estimation c l ng t ng ký hi u g n gi ng nh t MMSE Minimum Mean Square Error L i bình ph ti u MRC Maximum Ratio Combining K th pt s c cđ i MS Mobile Station Tr m di đ ng MTC CDMA Multi-code CDMA CDMA đa mư MTC-MC CDMA Multi-code Multicarrier CDMA CDMA đa sóng mang a mư MUI Multiple User Interference Nhi u đa ng Multicarrier-Code Division Multicarrier-CDMA Multiplexing cl ng tr c giao đa mư ng chu i g n gi ng nh t ng trung bình c c i dùng a truy nh p phân chia theo mư đa sóng mang OFDM Orthogonal Frequency Division Multiplexing Ghép kênh phân chia theo t n s tr c giao ORC Orthogonal Restoring Combing K t h p khôi ph c tr c giao P/S Parallel-to-Serial Song song sang n i ti p PG Processing Gain PIC Parallel Interference Cancellation Tri t nhi u song song PN Pseudo Noise Chu i gi ng u nhiên QPSK Quaternary Phase Shift Keying Khóa d ch pha c u ph S/P Serial-to-Parallel N i ti p sang song song l i x lý ng SINR Signal to Interference Noise Ratio T s công su t trung bình c a tín hi u t p âm c ng can nhi u SNR Signal-to-Noise (power) Ratio T s (công su t) tín hi u t p âm Danh m c hình v b ng Hình 1.1 Nguyên lý tr i ph theo DS-CDMA 11 Hình 1.2 a) Máy phát; b) Ph công su t phát; c) Máy thu RAKE 13 Hình 1.3 H th ng truy n d n đa sóng mang áp d ng OFDM 15 Hình 1.4 Mô hình h th ng MC CDMA 18 Hình 1.5 Nguyên lý tr i ph c a h th ng MC CDMA 18 Hình 1.6 S đ kh i h th ng Multicarrier-CDMA 20 Hình 1.7 S đ kh i h th ng MC CDMA v i N ≠ PG 21 Hình 1.8 B t o chu i Gold 26 Hình 1.9 S đ h th ng Multi-Code CDMA tr c giao thông th ng Hình 2.1 S đ c u trúc Máy phát (a) máy thu (b) c a h h ng TCMC CDMA Hình 2.2 K t qu mô ph ng c a BER theo SNR c a h MC-CDMA, MTC CDMA, MTC-MC CDMA v i M thay đ i T t c h th ng chi m t ng b ng t n nh Hình 2.3 So sánh gi a phân tích lý thuy t v i k t qu mô ph ng (L=16) c a BER theo SNR, tr ng h p kênh sóng mang đ c l p v i Hình 2.4: th BER theo s ng i s d ng h th ng MCCDMA h th ng MTC-MC CDMA Hình 2.5: SINR thu đ c Theo SNR v i K, M thay đ i Có th th y r ng giá tr M không đ i v i SINR phía thu Hinh 2.6 Minh h a kênh fading đa đ ng Hình 2.7 Minh h a đáp ng xung kim c a kênh biên d ng tr đa đ ng Hình 2.8 l i kênh fading Rayleigh 27 Hình 2.9 Biên d ng tr công su t c a mô hình tia thành ph 52 Hình 2.10 S u ki n t c đ thích nghi kênh fading đa đ ng Hình 2.11 th BER theo SNR c a h th ng MTC CDMA s d ng mư Walsh, Gold, Kasami kênh AWGN (S ng i dùng= 10 16) Hình 2.12 th BER theo S ng i s d ng h th ng MTC CDMA s d ng mư Walsh, Gold, Kasami kênh AWGN (SNR=10dB, M=16) Hình 2.13 th BER theo SNR h th ng MTC CDMA s d ng mư Walsh, Gold, Kasami kênh fading đa đ ng (s ng i dùng=10 16) Hình 2.14 th BER theo s ng i dùng h th ng MTC CDMA kênh AWGN (SNR=10dB) kênh fading (SNR=40dB), (M=16) 55 31 38 39 40 41 43 46 51 57 57 58 59 Hình 2.15 th BER theo s ng i dùng h th ng MTC CDMA v i có thích nghi t c đ kênh fading đa đ ng (SNR = 40dB, M=16 cho MTC CDMA không thích nghi t c đ ) Hình 2.16 th BER theo SNR cho MTC CDMA, MTC-MC CDMA thích nghi t c đ , MTC-MC CDMA có thích nghi t c đ kênh fading đa đ ng (s ng i s d ng =10, sóng mang con=16 cho MTC-MC, M=16 cho MTC CDMA MTC-MC CDMA không thích nghi t c đ ) Hình 2.17 Giá tr trung bình M theo s ng i dùng h th ng MTC-MC CDMA có thích nghi t c đ kênh fading đa đ ng (SNR=40dB) B ng 1.1 Các giá tr đ u si4 dùng ma tr n Hadamard 4×4 59 60 61 29 B ng 2.1 Biên d ng tr công su t c a d ng t bào đô th đô th x u v i 6-tia 51 B ng 2.2 Các thông s mô ph ng 56 L i nói đ u H th ng không dây t ng lai nh m ng t bào th h th (4G) đòi h i cung c p m t cách linh ho t cho thuê bao v i d ch v đa d ng nh tho i, d li u, nh truy n hình B i d ch v có t c đ d li u khác r t nhi u, h th ng th h t ng lai s ph i thích nghi v i tính đa d ng r t l n c a t c đ d li u a truy c p phân chia mư (CDMA) đư đ c ch ng minh r t thành công cho h th ng tho i t bào l n, nh ng có m t vài thái đ hoài nghi v vi c li u CDMA s thích h p t t v i truy n khác tho i i u đư thúc đ y nghiên c u h th ng CDMA đa mư cho phép t c đ d li u thay đ i b ng cách phân ph i đa mư, thay đ i dung l ng ng i s d ng khác Trong lúc CDMA đa sóng mang (MC-CDMA) đư n i lên nh nh s thay th m nh m CDMA chu i tr c ti p (DS-CDMA) truy n th ng thông tin vô n di đ ng, cho th y hi u qu t t h n CDMA đ n sóng mang kênh fading đa đ ng Trong lu n v n nêu phân tích k thu t u ch , đa truy c p m i s k t h p gi a h th ng CDMA đa sóng mang v i h th ng đa mư đ t n d ng l i th t t nh t c a m i h th ng tr c So v i k thu t truy n đa t c h th ng CDMA đ n sóng mang kênh AWGN đư đ c đ c p tr c đó, k thu t đa mư m i làm cân đ i gi a s thuê bao có th v i t c đ d li u m i thuê bao, nói cách khác s ng i s d ng đ ng th i t c đ cao h th ng CDMA đa mư s h n s ng i s d ng t c đ d li u b ng h th ng CDMA truy n th ng tr c Thay đ i s đ đa mư, s giúp thay đ i t c đ d li u b ng cách thay đ i t p h p chu i mư đ dùng Ng c ch đ nh cho m i ng i i dùng truy n d li u c a b ng cách ch n m t chu i t t h p mư c a h đ truy n kênh Nh v y, hi u n ng h CDMA đa mư ch coi nh kênh AWGN n đư có hai lo i nghiên c u truy n d n đa t c đ d a đa sóng mang–tr i ph chu i tr c ti p (MCDS CDMA) Th nh t nghiên c u truy n đa t c cho h th ng CDMA chu i tr c ti p đa sóng mang d a nguyên lý c b n g m truy c p đa mư th hai truy c p mư h s tr i bi n đ i Trong CDMA đa mư, dòng d li u c a ng i s d ng t c đ M ban đ u đ c ghép kênh vào M dòng liên ti p khác v i m t t c đ d li u c b n, m i dòng liên t c coi nh m t ng i s d ng riêng bi t Sau m i dòng liên ti p M đ thành P dòng nh song song đ c bi n đ i c tr i ph b i mư tr i ph gi ng m t h s tr i ph H th ng CDMA chu i tr c ti p đa sóng mang có M l n nhi u/ng đ c coi nh m t ng i s d ng, b i m i dòng M dòng d li u i dùng đ c l p Do đó, h th ng ch u nhi u nhi u h n t ng t c đ d li u, v i s ng i dùng c đ nh Trong h th ng CDMA đa sóng mang đa t c, sóng mang đ c phân chia vào M nhóm theo t c đ d li u yêu c u Do đó, s sóng mang c đ nh, h s tr i mi n t n s c a d li u b gi m t ng t c đ d li u H th ng CDMA đa mư đ t v n đ tr i r ng can nhi u b ng h đa mư b i ch s d ng m t chu i mư đ thay cho vi c tr i t ng dòng M dòng d li u đ c ghép kênh, làm cho nhi u không t ng n tính v i t c đ d li u Dù th h th ng CDMA đa mư c ng không th đ t đ c đ l i phân t p t n s c a u ch đa sóng mang H th ng CDMA đa sóng mang đa mư (MTC-MC CDMA ) s k t h p hoàn h o hai h th ng CDMA đa sóng mang h th ng CDMA đa mư, có đ c u m c a c hai h th ng nh : (i) t c đ d li u thay đ i mà không m r ng nhi u (ii) kh n ng ch ng kênh fading đa đ ng H n n a, h th ng có c h s tr i mi n t n s tr i mi n th i gian đ t n d ng hài hòa c phân t p trung bình hóa can nhi u c a đ c tính c a h th ng u ch đa sóng mang CDMA Hi u n ng c a h th ng MTC-MC CDMA đ c suy theo bi u th c gi i tích c i ti n h th ng t h th ng MC-CDMA, c ng thông qua nh ng phân tích đ c mô ph ng kênh fading ch n l c t n s N i dung c a nh ng ph n ti p theo c a lu n v n g m: Ph n I : Gi i thi u v h th ng CDMA lai ghép Ph nII : H th ng lai ghép CDMA đa sóng mang đa mư Ph l c: Mã ngu n m t vài ch ng trình ch y mô ph ng (Matlab)  MTC CDMA simulator in Multipath fading channel (mtcdma.m) function F = mtcdma(code,K,M,SNR,iteration) % Multi-code CDMA Simulator in Multipath fading channel Ver 2.0 by CDMA team % % Description : Simulator for Multi-cdoe CDMA Systems % (Using Multipath Rayleigh Fading channel % with "Typical macrocellular urban 6-ray power delay profiles" % from "Principles of MOBILE COMMUNICATION" by Gordon L Stuber) % % ***USAGE : % mtcdma(code,K,M,SNR,iteration),(e.g mtcdma(1,5,8,10,10000)) % X=mtcdma(code,K,M,SNR,iteration) returns a vector which is [Bit error rate,Symbol error rate] % % Parameter % code : = Walsh, = Gold, = Kasami, = M-sequence % K : No of Users % M : No of sequence which we use in the G(n) sequence set = M-ary % SNR : SNR (dB) % iteration : iteration number = symbol number % % Using m-files : bit_gen.m, m_seq_gen.m, gold.m, kasami.m, % BasicSeqK.m,BasicSeq.m, Gaussian.m, % "NEED CDMA Communication Toolbox" % FYI #1 : "00000" all zero bit is mapped to 1st symbol That means if the % binary value of information bits is n, then the symbol is (n+1)th % sequence % FYI #2 : In every vector [1, ,N], is the first bit and N is the % last bit in time domain and frequency domain % Generate sequence set G_set, G(user,time) G_set=[]; switch code case {1} G_set = hadamard(256); % Length of G = 256 case {2} G_set_temp = gold(1); % Length of G = 255, NOT 256 G_set = [G_set_temp ones(length(G_set_temp(:,1)),1)]; % Add "1" for the last bit(256th) to the code whose length is 255 case {3} G_set_temp = kasami(1); % Length of G = 255, NOT 256 G_set = [G_set_temp ones(length(G_set_temp(:,1)),1)]; % Add "1" for the last bit(256th) to the code whose length is 255 case {4} p = [1 0 1 1]; %(1) G_one = m_seq_gen(g,1); for I=1:255 G_set_temp(I,:) = circshift(G_one,[0,I-1]); end G_set(I,:) = [G_set_temp(I,:) 1]; % Add "1" for the last bit(256th) to the code whose length is 255 end % Generate user specific sequence U, U(user,time) p2 = [1 0 1 1]; %(2) g2 = m_seq_gen(p2,1); for I=1:K U_temp(I,:) = circshift(g2,[0,I-1]); U(I,:) = [U_temp(I,:) 1]; end E_stot = zeros(1,K); E_btot = 0; fd = 50; fs = 1.228*10^2; for i=1:4*K fad(i,:) = ray2(fd,fs,iteration/40,1); fad(i,:)= fad(i,:) / mean(fad(i,:)); end Resid = zeros(K,5); fad_a = [0.8070 0.095 0.061 0.037]; %fad_tau = [0 1.6 2.4 4.8]; % N_leng : Length of the sequence G(n) -> We fixed this length = 255 or 256 N_leng = 256; Z = zeros(K,N_leng); for itr = 1:iteration %*********************************** iteration = symbol number ********************************** % Generate the Tx bits & Make M-ary Symbol No_bit=log2(M); % Calculation the sigma for AWGN EbNo_temp = SNR/10.; N_temp = (N_leng)/(10^EbNo_temp); %%%%%%%%% for the same Es/No sigma = sqrt(N_temp/2); %dec=zeros(1,No_bit); symbol = bit_gen(1,No_bit,K); % symbol(1,time,user) symbol_temp = zeros(No_bit,K); symbol_temp(:,:) = symbol(1,:,:); symbol_index = bi2de(sign(symbol_temp+1)')+1; G = G_set(symbol_index,:); % Transmitter % Spread in Frequency domain, X:Spreaded Signal in Freq domain, X = G.*U; % Fading channel, output >SF() if itr == iteration fad_ind = floor(itr/40); else fad_ind = floor(itr/40) + 1; end Gain(:,1) = fad(:,fad_ind); %Gain(User*4,1), that means Gain(1:4,:) is for user and Gain(5:8,:) is for user SF = zeros(K,N_leng); i=1:K; SF(:,:) = X(i,:).*repmat(Gain(4*(i-1)+1,1),1,N_leng).*fad_a(1)+[Z(:,1:2) X(i,1:N_leng-2)].*repmat(Gain(4*(i-1)+2,1),1,N_leng).*fad_a(2)+[Z(:,1:3) X(i,1:N_leng-3)].*repmat(Gain(4*(i-1)+3,1),1,N_leng).*fad_a(3)+[Z(:,1:5) X(i,1:N_leng-5)].*repmat(Gain(4*(i-1)+4,1),1,N_leng).*fad_a(4); SF(:,:) = SF(:,:) + [Resid(i,4:5) Z(:,1:N_leng-2)].*fad_a(2).*repmat(Gain(4*(i1)+2,1),1,N_leng)+[Resid(i,3:5) Z(:,1:N_leng-3)].*fad_a(3).*repmat(Gain(4*(i1)+3,1),1,N_leng)+[Resid(i,:) Z(:,1:N_leng-5)].*fad_a(4).*repmat(Gain(4*(i1)+4,1),1,N_leng); Resid(:,:) = X(:,(N_leng-4):N_leng); SF_SUM = sum(SF,1); % AWGN, R(chip index) R = SF_SUM + sigma*randn(1,N_leng); % Receiver, R_desp(K,) R_desp = repmat(R,K,1).*U; % Detection using Matched filter for G(n), out_mat(user,output value of matched filter for % each G(n) sequence) out_mat = R_desp*G_set'; % Decode : each index = user No , each elements = Decimal value of maximum output sequence of % matched value [maximum,Decode] = max((out_mat(:,1:M))'); result = Decode - symbol_index'; % Calculate the symbol error E_symbol = sign(abs(result)); E_stot = E_stot + E_symbol; % Calculate the bit error Decode_t = (Decode - 1)'; E_bit_temp = sign(de2bi(Decode_t,No_bit)-0.5) - symbol_temp'; E_bit = sign(abs(E_bit_temp)); E_btot = E_btot + sum(sum(E_bit,1),2); end E_avg = sum(E_stot,2); SER = E_avg/K/iteration; BER = E_btot/K/iteration/No_bit; F = [BER, SER];  MTC-MC CDMA simulator in Multipath fading channel (mccdma.m) function F=mccdma(code,K,M,Sub,SNR,iteration) % Multi-code Multi-Carrier CDMA Simulator in Multipath fading channel % % Description % Parameter % K : No of Users % M : No of sequence which we use in the G(n) sequence set = M-ary % N : Length of the sequence G(n) -> We fixed this length = 255 or 256 % Sub : Number of Subcarrier frequency (Sub SHOULD BE or 16) % % Using m-files : bit_gen.m, m_seq_gen.m % % ***USAGE : mccdma(code,K,M,Sub,SNR,iteration), (e.g mccdma(1,5,32,16,10,10000)) % X=mccdma(code,K,M,Sub,SNR,iteration) returns a vector which is [Bit error rate, Symbol % error rate] % % FYI #1 : "00000" all zero bit is mapped to 1st symbol That means if the % binary value of information bits is n, then the symbol is (n+1)th % sequence % FYI #2 : In every vector [1, ,N], is the first bit and N is the % last bit in time domain and frequency domain % Define the value for parameters N = 256; % Generate sequence set G_set, G(user,time) G_set=[]; switch code case {1} G_set = hadamard(256); % Length of G = 256 case {2} G_set_temp = gold(1); % Length of G = 255, NOT 256 G_set = [G_set_temp ones(length(G_set_temp(:,1)),1)]; % Add "1" for the last bit(256th) to the code whose length is 255 case {3} G_set_temp = kasami(1); % Length of G = 255, NOT 256 G_set = [G_set_temp ones(length(G_set_temp(:,1)),1)]; % Add "1" for the last bit(256th) to the code whose length is 255 case {4} p = [1 0 1 1]; %(1) G_one = m_seq_gen(g,1); for I=1:255 G_set_temp(I,:) = circshift(G_one,[0,I-1]); end G_set(I,:) = [G_set_temp(I,:) 1]; % Add "1" for the last bit(256th) to the code whose length is 255 end % Generate user specific sequence U, U(user,time) switch Sub case {8} p2 = [1 1]; otherwise p2 = [1 0 1]; Sub = 16; end g2 = m_seq_gen(p2,1); for I=1:K U_temp(I,:) = circshift(g2,[0,I-1]); U(I,:) = [U_temp(I,:) 1]; end %U = sign(rand(K,Sub)-.5); U_t=[]; for i=1:K U_t=cat(1,U_t,repmat(U(i,:),N,1)); end E_stot = zeros(1,K); E_btot = 0; fd = 50; fs = 1.228*10^2; for i=1:4*K fad(i,:) = ray2(fd,fs,iteration/40,1); fad(i,:)= fad(i,:) / mean(fad(i,:)); end Resid = zeros(K,80); fad_a = [0.8070 0.095 0.061 0.037]; N = 256; Z = zeros(K,N*Sub); symbol_T = N*Sub; E_btot = 0; for itr = 1:iteration %*********************************** iteration = symbol number ********************************** % Generate the Tx bits & Make M-ary Symbol No_bit=log2(M); % Calculation the sigma for AWGN EbNo_temp = SNR/10.; N_temp = (N*Sub)/(10^EbNo_temp); %%%%%%%%% sigma = sqrt(N_temp/2); symbol = bit_gen(1,No_bit,K); % symbol(1,time,user) symbol_temp = zeros(No_bit,K); symbol_temp(:,:) = symbol(1,:,:); symbol_index = bi2de(sign(symbol_temp+1)')+1; % Select G(n) for each user depending on the information bit(M-ary symbol), % G(user,time) G = G_set(symbol_index,:); if itr == iteration fad_ind = floor(itr/40); else fad_ind = floor(itr/40) + 1; end Gain(:,1) = fad(:,fad_ind); %Gain(User*4,1), that means Gain(1:4,:) is for user and Gain(5:8,:) is for user % Transmitter % Spread in Frequency domain, X:Spreaded Signal in Freq domain, X = repmat(reshape(G',prod(size(G)),1),1,Sub).*U_t; % Do IFFT operation across the rows (dimension = 2) to code into time, S = sqrt(Sub).*ifft(X,Sub,2); S_t = reshape(S',symbol_T,K)'; % Fading channel, output >SF() % ************ We should put the fading channel part at here ******** SF = zeros(K,symbol_T); i=1:K; SF(:,:) = S_t(i,:).*repmat(Gain(4*(i-1)+1,1),1,symbol_T).*fad_a(1)+[Z(:,1:2) S_t(i,1:symbol_T-2)].*repmat(Gain(4*(i-1)+2,1),1,symbol_T).*fad_a(2)+[Z(:,1:3) S_t(i,1:symbol_T-3)].*repmat(Gain(4*(i-1)+3,1),1,symbol_T).*fad_a(3)+[Z(:,1:5) S_t(i,1:symbol_T-5)].*repmat(Gain(4*(i-1)+4,1),1,symbol_T).*fad_a(4); SF(:,:) = SF(:,:) + [Resid(i,64:80) Z(:,1:symbol_T-17)].*fad_a(2).*repmat(Gain(4*(i1)+2,1),1,symbol_T)+[Resid(i,48:80) Z(:,1:symbol_T33)].*fad_a(3).*repmat(Gain(4*(i-1)+3,1),1,symbol_T)+[Resid(i,:) Z(:,1:symbol_T80)].*fad_a(4).*repmat(Gain(4*(i-1)+4,1),1,symbol_T); Resid(:,:) = S_t(:,(symbol_T-79):symbol_T); SF_SUM = sum(SF,1); % AWGN, R(1,N*Sub) R_t = SF_SUM + sigma*randn(1,symbol_T) + j*sigma*randn(1,symbol_T); R = reshape(R_t',Sub,N)'; % Receiver, R(sequence time index, sub-carrier index) R_fft = 1/sqrt(Sub)*fft(R,Sub,2); % Despread, R_desp(user,sequence time index) for i=1:K R_desp_temp(i,:) = (R_fft*U(i,:)'/Sub)'; end R_desp = real(R_desp_temp); out_mat = R_desp*G_set(1:M,:)'; [maximum,Decode] = max(out_mat'); %Decoded_bits = G_set(Decode,:); result = Decode - symbol_index'; % Calculate the symbol error E_symbol = sign(abs(result)); E_stot = E_stot + E_symbol; % Calculate the bit error Decode_t = (Decode - 1)'; E_bit_temp = sign(de2bi(Decode_t,No_bit)-0.5) - symbol_temp'; E_bit = sign(abs(E_bit_temp)); E_btot = E_btot + sum(sum(E_bit,1),2); end E_avg = sum(E_stot,2); SER = E_avg/K/iteration; BER = E_btot/K/iteration/No_bit; F = [BER, SER];  Adaptive function for MTC-MC CDMA system (AdapRateMC.m) function M = AdaptRateMC(R, K, sigma2, SG, MaxM, Sub) % Rate Calculation for Multicode Multicarrier CDMA Simulator v 1.5 by CDMA team % % usage: [M] = AdaptRate(r, sigma, SG) % R : Channel Gain of each user, [R] = r(1 K*4) % K : No of Users % sigma2 : AWGN power = No/2 % SG : Spreading Gain % MaxM : Maximum allowable Rate M % % [M] : Tx Rate of Each User K MinM = 2; i = 0:K-1; r(i+1) = (0.8070*R(4*i+1) + 0.095*R(4*i+2) + 0.061*R(4*i+3) + 0.037*R(4*i+4)); % averaging each user's channel gain rs = r*r'; % total recieved power ri = r.*r; % signal power gamma = ri./(rs - ri + sigma2)*SG*Sub; Pb = 10^-3; K = * log(Pb/0.1); M = 2.^(-1*floor(gamma./K/3)); M = min(max(M,MinM), MaxM);  MTC-MC CDMA simulator with adaptation in Multipath fading channel (mcAdap.m) function F=mcAdap(code,K,MaxM,Sub,SNR,iteration) % Multi-code Multi-Carrier CDMA Simulator with Adaptive algorithm in % Multifath fading channel % Description % Parameter % K : No of Users % M : No of sequence which we use in the G(n) sequence set = M-ary % N : Length of the sequence G(n) -> We fixed this length = 255 or 256 % Sub : Number of Subcarrier frequency (Sub SHOULD BE or 16) % Using m-files : bit_gen.m, m_seq_gen.m % % ***USAGE : mcAdap(c,K,MaxM,Sub,SNR,iteration), (e.g mcAdap(1,5,32,16,10,10000)) % X=mcAdap(c,K,MaxM,Sub,SNR,iteration) returns a vector which is [Bit error rate, Symbol error rate] % % FYI #1 : "00000" all zero bit is mapped to 1st symbol That means if the % binary value of information bits is n, then the symbol is (n+1)th % sequence % FYI #2 : In every vector [1, ,N], is the first bit and N is the % last bit in time domain and frequency domain % Define the value for parameters N = 256; % Generate sequence set G_set, G(user,time) G_set=[]; switch code case {1} G_set = hadamard(256); % Length of G = 256 case {2} G_set_temp = gold(1); % Length of G = 255, NOT 256 G_set = [G_set_temp ones(length(G_set_temp(:,1)),1)]; % Add "1" for the last bit(256th) to the code whose length is 255 case {3} G_set_temp = kasami(1); % Length of G = 255, NOT 256 G_set = [G_set_temp ones(length(G_set_temp(:,1)),1)]; % Add "1" for the last bit(256th) to the code whose length is 255 case {4} p = [1 0 1 1]; %(1) G_one = m_seq_gen(g,1); for I=1:255 G_set_temp(I,:) = circshift(G_one,[0,I-1]); end G_set(I,:) = [G_set_temp(I,:) 1]; % Add "1" for the last bit(256th) to the code whose length is 255 end % Generate user specific sequence U, U(user,time) switch Sub case {8} p2 = [1 1]; otherwise p2 = [1 0 1]; Sub = 16; end g2 = m_seq_gen(p2,1); for I=1:K U_temp(I,:) = circshift(g2,[0,I-1]); U(I,:) = [U_temp(I,:) 1]; end U_t=[]; for i=1:K U_t=cat(1,U_t,repmat(U(i,:),N,1)); end E_stot = zeros(1,K); E_btot = 0; fd = 50; fs = 1.228*10^2; for i=1:4*K fad(i,:) = ray2(fd,fs,iteration/40,1); fad(i,:)= fad(i,:) / mean(fad(i,:)); end Resid = zeros(K,80); fad_a = [0.8070 0.095 0.061 0.037]; N = 256; Z = zeros(K,N*Sub); symbol_T = N*Sub; M = 2*ones(1,K); E_btot = zeros(1,K); No_bit_tot = zeros(1,K); kjw = zeros(iteration,K); kty = zeros(iteration,K); for itr = 1:iteration %*********************************** iteration = symbol number ********************************** % Generate the Tx bits & Make M-ary Symbol No_bit=log2(M); % Calculation the sigma for AWGN EbNo_temp = SNR/10.; N_temp = (N*Sub)/(10^EbNo_temp); %%%%%%%%% sigma = sqrt(N_temp/2); if itr == iteration fad_ind = floor(itr/40); else fad_ind = floor(itr/40) + 1; end Gain(:,1) = fad(:,fad_ind); %Gain(User*4,1), that means Gain(1:4,:) is for user and Gain(5:8,:) is for user for i=1:K symbol = bit_gen(1,No_bit(i),1); % symbol(1,time,user) symbol_index(i) = bi2de(sign(symbol+1))+1; G(i,:) = G_set(symbol_index(i),:); end % Transmitter % Spread in Frequency domain, X:Spreaded Signal in Freq domain, X = repmat(reshape(G',prod(size(G)),1),1,Sub).*U_t; % Do IFFT operation across the rows (dimension = 2) to code into time, S = sqrt(Sub).*ifft(X,Sub,2); S_t = reshape(S',symbol_T,K)'; SF = zeros(K,symbol_T); i=1:K; SF(:,:) = S_t(i,:).*repmat(Gain(4*(i-1)+1,1),1,symbol_T).*fad_a(1)+[Z(:,1:2) S_t(i,1:symbol_T-2)].*repmat(Gain(4*(i-1)+2,1),1,symbol_T).*fad_a(2)+[Z(:,1:3) S_t(i,1:symbol_T-3)].*repmat(Gain(4*(i-1)+3,1),1,symbol_T).*fad_a(3)+[Z(:,1:5) S_t(i,1:symbol_T-5)].*repmat(Gain(4*(i-1)+4,1),1,symbol_T).*fad_a(4); SF(:,:) = SF(:,:) + [Resid(i,64:80) Z(:,1:symbol_T-17)].*fad_a(2).*repmat(Gain(4*(i1)+2,1),1,symbol_T)+[Resid(i,48:80) Z(:,1:symbol_T33)].*fad_a(3).*repmat(Gain(4*(i-1)+3,1),1,symbol_T)+[Resid(i,:) Z(:,1:symbol_T80)].*fad_a(4).*repmat(Gain(4*(i-1)+4,1),1,symbol_T); Resid(:,:) = S_t(:,(symbol_T-79):symbol_T); SF_SUM = sum(SF,1); % AWGN, R(1,N*Sub) R_t = SF_SUM + sigma*randn(1,symbol_T) + j*sigma*randn(1,symbol_T); R = reshape(R_t',Sub,N)'; % Receiver, R(sequence time index, sub-carrier index) R_fft = 1/sqrt(Sub)*fft(R,Sub,2); % Despread, R_desp(user,sequence time index) for i=1:K R_desp_temp(i,:) = (R_fft*U(i,:)'/Sub)'; end R_desp = real(R_desp_temp); % Detection using Matched filter for G(n), out_mat(user,output value of matched filter for % each G(n) sequence) % Decode : each index = user No , each elements = Decimal value of maximum output sequence of % matched value for i=1:K out_mat = R_desp(i,:)*G_set(1:M(i),:)'; [maximum,Deco] = max(out_mat); Decode(i) = Deco; end %Decoded_bits = G_set(Decode,:); result = Decode - symbol_index; % Calculate the symbol error E_symbol = sign(abs(result)); E_stot = E_stot + E_symbol; % Calculate the bit error for i=1:K Decode_t = (Decode(i) - 1)'; E_bit_temp = sign(de2bi(Decode_t,No_bit(i))-0.5) - sign(de2bi((symbol_index(i)1)',No_bit(i))-0.5);%symbol_temp'; E_bit = sign(abs(E_bit_temp)); E_btot(i) = E_btot(i) + sum(E_bit); No_bit_tot(i) = No_bit_tot(i) + No_bit(i); end %%% Calculate Tx Rate M here M = AdaptRateMC(mean(Gain,2), K, N_temp/2, N, MaxM, Sub); kty(itr,:) = M; i=0:K-1; kjw(itr,i+1) = (0.8070*mean(Gain(4*i+1),2) + 0.095*mean(Gain(4*i+2),2) + 0.061*mean(Gain(4*i+3),2) + 0.037*mean(Gain(4*i+4),2))'; end SER = sum(E_stot,2)/K/iteration; BER = sum(E_btot)/sum(No_bit_tot); BER_user = E_btot./No_bit_tot; Mean_M = mean(mean(kty)); F = [BER, Mean_M, SER, BER_user]  H2_11.m clc, figure, i=6:15; inter=1000; % F = mtcdma(code,K,M,SNR,iteration) % Parameter % code : = Walsh, = Gold, = Kasami, = M-sequence % K : No of Users % M : No of sequence which we use in the G(n) sequence set = M-ary % SNR : SNR (dB) % iteration : iteration number = symbol number(=10000) disp('- Tinh BER cua ma Walsh:'), f_ber1=zeros(1,length(i)); for k=1:10 f=mtawgn(1,10,16,i(k),inter); f_ber1(k)=f(1) end semilogy(i,f_ber1,'b-o') hold on, grid on % disp('- Tinh BER cua ma Gold:'), f_ber2=zeros(1,length(i)); for k=1:10 f=mtawgn(2,10,16,i(k),inter); f_ber2(k)=f(1) end semilogy(i,f_ber2,'g-s') hold on, % disp('- Tinh BER cua ma Kasami:'), f_ber3=zeros(1,length(i)); for k=1:10 f=mtawgn(3,10,16,i(k),inter); f_ber3(k)=f(1) end semilogy(i,f_ber3,'r-*') semilogy([8],[.12],'w-'), semilogy([9],[.123],'w-') % dung de tao hai dong chu thich xlabel('SNR (dB)'),ylabel('BER') axis([6 15 001 1]) legend('Walsh code','Gold code','Kasami code','No.of User=10','AWGN')  h2_12.m clc, figure, i=[1 10 15 20 25 30 40 50]; inter=1000; % F = mtcdma(code,K,M,SNR,iteration) % Parameter % code : = Walsh, = Gold, = Kasami, = M-sequence % K : No of Users % M : No of sequence which we use in the G(n) sequence set = M-ary % SNR : SNR (dB) % iteration : iteration number = symbol number(=10000) disp('- Tinh BER cua ma Walsh:'), f_ber1=zeros(1,length(i)); for k=1:11 f=mtawgn(1,i(k),16,10,inter); f_ber1(k)=f(1) end semilogy(i,f_ber1,'b-o') hold on, grid on % disp('- Tinh BER cua ma Gold:'), f_ber2=zeros(1,length(i)); for k=1:11 f=mtawgn(2,i(k),16,10,inter); f_ber2(k)=f(1) end semilogy(i,f_ber2,'g-s') hold on, % disp('- Tinh BER cua ma Kasami:'), f_ber3=zeros(1,length(i)); for k=1:11 f=mtawgn(3,i(k),16,10,inter); f_ber3(k)=f(1) end semilogy(i,f_ber3,'r-*') semilogy([8],[.12],'w-'), semilogy([9],[.123],'w-') % dung de tao hai dong chu thich xlabel('No.of Users'),ylabel('BER') axis([0 50 001 1]) legend('Walsh code','Gold code','Kasami code','SNR=10dB','AWGN',4)  h2_13.m clc, figure, i=5:5:40; inter=1000; % F = mtcdma(code,K,M,SNR,iteration) % Parameter % code : = Walsh, = Gold, = Kasami, = M-sequence % K : No of Users % M : No of sequence which we use in the G(n) sequence set = M-ary % SNR : SNR (dB) % iteration : iteration number = symbol number(=10000) disp('- Tinh BER cua ma Walsh:'), f_ber1=zeros(1,length(i)); for k=1:8 f=mtcdma(1,10,16,i(k),inter); f_ber1(k)=f(1) end semilogy(i,f_ber1,'b-o') hold on, grid on % disp('- Tinh BER cua ma Gold:'), f_ber2=zeros(1,length(i)); for k=1:8 f=mtcdma(2,10,16,i(k),inter); f_ber2(k)=f(1) end semilogy(i,f_ber2,'g-s') hold on, % disp('- Tinh BER cua ma Kasami:'), f_ber3=zeros(1,length(i)); for k=1:8 f=mtcdma(3,10,16,i(k),inter); f_ber3(k)=f(1) end semilogy(i,f_ber3,'r-*') semilogy([8],[.12],'w-'), semilogy([9],[.123],'w-') % dung de tao hai dong chu thich xlabel('SNR (dB)'),ylabel('BER') axis([5 40 01 1]) legend('Walsh code','Gold code','Kasami code','No.of User=10','Multi-path Fading')  h2_14.m clc, home, figure, i=[3 10 15]; inter=1000; % F = mtcAdap(code,K,MaxM,SNR,iteration) % Parameter % code : = Walsh, = Gold, = Kasami, = M-sequence % K : No of Users % MaxM : Max No of sequence which we use in the G(n) sequence set = M-ary % SNR : SNR (dB) % iteration : iteration number = symbol number % disp('- Tinh BER cua MTC CDMA co thich nghi toc do:'), f_ber1=zeros(1,length(i)); for k=1:5 f=mtcAdap(1,i(k),16,40,inter); f_ber1(k)=f(2) end semilogy(i,f_ber1,'b-o') hold on, grid on % disp('- Tinh BER cua MTC CDMA khong thich nghi toc do:'), f_ber2=zeros(1,length(i)); for k=1:5 f=mtcdma(1,i(k),16,40,inter); f_ber2(k)=f(1) end semilogy(i,f_ber2,'g-s') hold on, % semilogy([8],[.01],'w-') % semilogy([9],[.123],'w-') dung de tao hai dong chu thich xlabel('No.of Users'),ylabel('BER') axis([3 15 0004 1]) legend('MTC CDMA with Adaptation','MTC CDMA without Adaptation','SNR=40dB',4)  h2_15.m clc, figure, i=[3 10 15]; inter=2000; % F = mtcAdap(code,K,MaxM,SNR,iteration) % Parameter % code : = Walsh, = Gold, = Kasami, = M-sequence % K : No of Users % MaxM : Max No of sequence which we use in the G(n) sequence set = M-ary % SNR : SNR (dB) % iteration : iteration number = symbol number % disp('- Tinh BER cua MTC CDMA kenh AWGN:'), f_ber1=zeros(1,length(i)); for k=1:5 f=mtawgn(1,i(k),16,10,inter); f_ber1(k)=f(1) end semilogy(i,f_ber1,'b-o') hold on, grid on % disp('- Tinh BER cua MTC CDMA kenh fading:'), f_ber2=zeros(1,length(i)); for k=1:5 f=mtcdma(1,i(k),16,40,inter); f_ber2(k)=f(1) end semilogy(i,f_ber2,'g-s') hold on, % semilogy([8],[.012],'w-') % semilogy([9],[.123],'w-') dung de tao hai dong chu thich xlabel('No.of Users'),ylabel('BER') axis([3 15 001 2]) legend('MTC CDMA in AWGN(SNR=10dB)','MTC CDMA in Fading(SNR=40dB)','M=16',4)  h2_16.m clc, figure, i=5:5:45; inter=1000; % -disp('- Tinh BER cua MTC CDMA khong co thich nghi toc do:'), f_ber1=zeros(1,length(i)); for k=1:9 f=mtcdma(1,10,16,i(k),inter); f_ber1(k)=f(1) end semilogy(i,f_ber1,'b-o') hold on, grid on % disp('- Tinh BER cua MTC-MC CDMA khong co thich nghi toc do:'), f_ber2=zeros(1,length(i)); for k=1:9 f=mccdma(1,10,16,16,i(k),inter); f_ber2(k)=f(1) end semilogy(i,f_ber2,'g-*') hold on, % disp('- Tinh BER cua MTC-MC CDMA co thich nghi toc do:'), f_ber3=zeros(1,length(i)); for k=1:9 f=mcAdap(1,10,32,16,i(k),inter); f_ber3(k)=f(1) end semilogy(i,f_ber3,'r-s') % semilogy([8],[.12],'w-') % semilogy([9],[.123],'w-') dung de tao hai dong chu thich xlabel('SNR (dB)'),ylabel('BER') axis([5 45 00007 5]) legend('MTC CDMA without Adaptation','MTC-MC CDMA without Adaptation','MTC-MC CDMA with Adaptation') ... i thi u v h th ng CDMA lai ghép Ph nII : H th ng lai ghép CDMA đa sóng mang đa mư Ph n I Gi i thi u h th ng CDMA lai ghép 1.1 K thu t tr i ph CDMA [9] [10] K thu t tr i ph CDMA, v lý thuy t b... hình h th ng CDMA đa mã Ph n II: H th ng lai ghép CDMA đa sóng mang đa mã 27 30 2.1 Mô hình h th ng MTC-MC CDMA 30 2.2 Phân tích h th ng MTC-MC CDMA 33 2.3 Hi u n ng c a h th ng MTC-MC CDMA 38 2.4... thi u H th ng CDMA lai ghép 11 1.1 K thu t tr i ph CDMA 11 1.2 H th ng CDMA đa sóng mang 13 1.2.1 Mô t h th ng truy n d n a sóng mang 13 1.2.2 Mô hình h th ng lai ghép CDMA đa sóng mang 17 1.2.3