Blind Identification of possibly under-determined convolutive MIMO systems
Communications and Signal Processing Laboratory
ECE Department, Drexel University
Philadelphia, PA 19104, USA
Oct, 2007
http://www.ece.drexel.edu/CSPL
______________________________________________________________________________
Introduction:
This package contains the MATLAB code for the algorithm proposed in the following
papers. The goal of this Matlab script is to identify a possibly under-determined convolutive MIMO system with multiple independent inputs and additive Gaussian noise.
This code will plot the blind estimation results using figures.
This folder contains the following matlab functions:
ISPDmain, is the main function for the 3rd order ISPD method, which runs Monte-Carlo runs for different channels, as well as different SNR.
ISPD4th, is the main function for the 4th order ISPD method, which runs Monte-Carlo runs for different channels, as well as different SNR.
1.function [A,B,C,FIT,IT]=comfac(X,Fac,Options,DoComp,CompIt,Init);
%COMFAC Algorithm for fitting the complex-valued PARAFAC model
The code is downloaded from http://www.models.kvl.dk/users/rasmus/presentations/comfac/index.htm
2. function [cum] = g_cc4_3D(signal_1,signal_2,signal_3,signal_4,C_LENGTH)
% This function generate the fourth order cross cumulant of 4 signals
The code is wrote by Binning Chen, avalable at http://www.ece.drexel.edu/CSPL
3. function y_cum = cum4x (w,x,y,z, maxlag, nsamp, overlap, flag, k1, k2)
%CUM4X Fourth-order cross-cumulants
The code is wrote by A. Swami
4. function [bis, bis_true, bis_true1]=estCum1(NF,Le,x,h,H)
% This function only calculates the true and estimated
% third order cumulant and bispectrum
5. function [Tris,Tris_true]=estTris4(NF,Le,k2,k3,x,n,h,H)
% This function only calculates the true and estimated
% Fourth order cumulant and tripectrum
6.function [h_est]=ComputeISPD4th(real_channel,draw_fig,NF,L_diff,k2,k3,Tris,h,H,N)
%%%%% This function calculate the estimated MIMO channel gain for given
%%%%% Trispectrum and other coefficients
7. function [h_est]=ComputeISPD3rd(real_channel,draw_fig,NF,L_diff,Bis,h,H,N,delta,R)
%%%%% This function calculate the estimated MIMO channel gain for given
%%%%% Bispectrum and other coefficients
8. function [AkrpB] = krp(A,B);
%%%% calculate the khatri-rao product of A and B
9. function y = rcos(x,alpha,T);
% raised cosine pulse genaration
10.function q=QAM(N,L,seed1,seed2)
% FUNCTION q = QAM(N,L);
% -----------------
% Generates L-QAM complex sequences Xr(n)+jXi(n)
11.function [PERM]=allign3(estC2,estC0,n)
%%% solve for the permutation ambiguioty between two PARAFAC iteration
12.function [H_est_even,h_est]=relign4(N, F, h, real_channel, L_diff, delta)
%%%% Recover the time domain channel from the frequency domain estimates
13. function [cum3_true] = True_cum3(h1,h2,h3,cum_length)
% True_cum3.M
% Calculate the True third order cumulant of three system driven by the same
% white input.
The code is wrote by Binning Chen, avalable at http://www.ece.drexel.edu/CSPL
14.function [cum4_3_true] = true_cum4_3D(h1,h2,h3,h4,cum_length)
% True_cum4_3D.M
% Calculate the True fourth order cumulant of three system driven by the same
% white input.
The code is wrote by Binning Chen, avalable at http://www.ece.drexel.edu/CSPL
15. function [cx] = g_cc3_matrix(signal_1,signal_2,signal_3,C_LENGTH)
% g_cc3_matrix.m
% This function generate the third order cross cumulant of 3 signals
The code is wrote by Binning Chen, avalable at http://www.ece.drexel.edu/CSPL
16.function [A,B,C,D]=QALS(X,F,A,B,C,D);
% Plain Vanilla Quadrilinear ALS
% X = IxJxKxL rank-F 4-way array
% Nikos Sidiropoulos
17.function y_cum = cum2x (x,y, maxlag, nsamp, overlap, flag)
%CUM2X Cross-covariance
The code is wrote by A. Swami
----------------------------------------------------------------------------------
Last Updated: Thursday, Oct 4, 2007
1. Y. Yu, and A.P. Petropulu, ※PARAFAC Based Blind Estimation Of Possibly Under-determined Convolutive MIMO Systems,§ IEEE Trans. on Signal Processing, accepted in 2007
2. T. Acar, Y. Yu, and A.P. Petropulu, ※Blind MIMO system estimation based on PARAFAC decomposition of tensors formed based HOS of the system output,§ IEEE Trans. on Signal Processing, Vol. 54, Nov. 2006.
3. Y. Yu, and A.P. Petropulu; ※Reduced Complexity Blind Estimation of Under-Determined Convolutive MIMO Systems§ IEEE 12th Digital Signal Processing Workshop Page(s):239 每 244, Sep. 2006
4. Y. Yu, and A.P. Petropulu, ※PARAFAC Based Blind Estimation of MIMO Systems with possibly more inputs than outputs§, IEEE International Conf on Acoustics, Speech and Signal Processing, Toulouse, France, May, 2006.
5. Y. Yu and A.P. Petropulu, ※Blind Estimation Of A Class Of Under-Determined Convolutive MIMO Systems Using Parafac Decomposition Of Output Tensors§, IEEE 40th Annual Conference on Information Sciences and Systems, Princeton, NJ, March 2006.
6. Y. Yu and A.P. Petropulu, ※Improved Single PARAFAC Decomposition based Blind MIMO Systems Estimation§, The 5th IEEE Symposium on Signal Processing and Information Technology, Athens, Greece, December, 2005
7. Y. Yu, and A.P. Petropulu, ※Improved PARAFAC based Blind MIMO system estimation,§ IEEE Thirty-Ninth Annual Asilomar Conference on Signal, Systems, and Computers, Pacific Grove, CA, USA, Oct 2005
8. Y. Yu, and A.P. Petropulu, ※Blind MIMO system estimation based on PARAFAC decomposition of HOS tensors,§ IEEE Workshop on Statistical Signal Processing, Bordeaux, France, July,2005.
评论3