%PERFORM Evaluate the performance of an TEQ in terms of% SSNR, SNR, geometric SNR, and channel capacity.% [SS, S, Si, Sg, Mg, Bf, Bm, Rf, Rm, Hw, Fh, Fw, Nc, Fhw] = % PERFORM(W,B,H,D,Nb,NN,X,N,Ph,Px,Pn,M,C,Fs,Mi) returns% the shortening SNR in SS, the SNR at the output of the % equalizer in S, the SNR distribution over the subchannels% in the vector Si. Sg is the real geometric SNR achieved% with the TEQ and Mg is the geometric SNR that can be % achieved in the case of zero ISI (Mg is calculated using% the MFB distribution while Sg is calculated with the SNR% distribution.). Bf is the number of bits per symbol % achievable with the TEQ and Bm is the upperbound on% bits per symbol. Rf is the channel capacity achieved with % the TEQ and Rm is the upperbound on the channel capacity.% Hw is a vector containing the equalized channel impulse % response. Fh is a vector containing the frequency response % of the original channel and Fw is vector containing the % equalizer. Nc is a vector of the power spectrum of the % channel noise after equalization and Fhw is a vector % containing the frequency response of the equalized channel.%% W is TEQ impulse response. B is the target impulse response% (for MMSE based techniques). H is the channel impulse% response. D is the optimal delay with TEQ. Nb is the target% window size. NN is the FFT size used in the DMT modulation.% X is the transmitted signal. N the channel noise. Ph is the % magnitude square of the channel frequency response. % Px is the power spectrum of the transmitted signal. Pn is the % power spectrum of the channel noise. M is the desired system % margin in dB which is used for channel capacity calculations.% C is the coding gain in dB assumed for channel capacity % calculations. Fs is the sampling frequency. Mi is the matched% filter bound distribution over frequency.% Copyright (c) 1999-2002 The University of Texas% All Rights Reserved.%  % This program is free software; you can redistribute it and/or modify% it under the terms of the GNU General Public License as published by% the Free Software Foundation; either version 2 of the License, or% (at your option) any later version.%  % This program is distributed in the hope that it will be useful,% but WITHOUT ANY WARRANTY; without even the implied warranty of% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the% GNU General Public License for more details.%  % The GNU Public License is available in the file LICENSE, or you% can write to the Free Software Foundation, Inc., 59 Temple Place -% Suite 330, Boston, MA 02111-1307, USA, or you can find it on the% World Wide Web at http://www.fsf.org.%  % Programmers:	Guner Arslan% Version:      @(#)perform.m	1.3  09/25/00% % The authors are with the Department of Electrical and Computer% Engineering, The University of Texas at Austin, Austin, TX.% They can be reached at arslan@ece.utexas.edu.% Guner Arslan is also with the Embedded Signal Processing% Laboratory in the Dept. of ECE., http://anchovy.ece.utexas.edu.function [SSNR, SNR, SNRi,geoSNRfinal,geoSNRmfb,bDMTfinal,bDMTmfb,...          RDMTfinal,RDMTmfb,hw,Fh,Fw,colorNoiseaft,Shw] = ...   perform(W,B,channel,D,Nb,N,inputSignal,noise,...           channelGain,inputSpec,noiseSpec,margin,codingGain,fs,MFBi,used)%W = W/norm(W);% equalized channel hw = conv(W,channel);%hw = hw(1:N);%hw = conv(W,channel);%hw = hw(1:N);% target windowwin = zeros(length(hw),1); win(D+1:D+Nb) = ones(1,Nb); % signal path impulse responsehwin = hw.*win;hwin = hwin - mean(hwin);% ISI path impulse responsehwout = hw.*(1-win);     hwout = hwout - mean(hwout);% energy of the ISI path normhwout = norm(hwout,2)^2;% remaining energy is signal normhwin = norm(hwin,2)^2;   % signal path energy to ISI energy gives the SSNRSSNR = 10*log10(normhwin./normhwout); % received signal power after equalizerfy = cov(filter(hw,1,inputSignal));  %FIXED: W->hw% received noise power after equalizerfn = cov(filter(W,1,noise));  % SNR at the receiver at the output of the equalizerSNR = 10*log10(fy./fn);   % equivalent path frequency responses in the used subchannelsFhwu = fft(hw,N)*sqrt(N); Fhw = Fhwu(1:N/2+1).'; %Fhw = Fhw(used);Fhu = fft(channel,N)*sqrt(N); Fh = Fhu(1:N/2+1).'; %Fh = Fh(used);% TIR frequency responseFbu = fft([zeros(D,1); B],N)*sqrt(N); Fb = Fbu(1:N/2+1).'; %Fb = Fb(used);% TEQ frequency responseFwu = fft(W,N)*sqrt(N); Fw = Fwu(1:N/2+1).';  %Fw = Fw(used);Fhwinu = fft(hwin,N)*sqrt(N); Fhwin = Fhwinu(1:N/2+1).'; %Fhwin = Fhwin(used);Fhwoutu = fft(hwout,N)*sqrt(N); Fhwout = Fhwoutu(1:N/2+1).'; %Fhwout = Fhwout(used);Shw = channelGain.*abs(Fw).^2;%Fhwin = sqrt(Shw);% noise power spectrumcolorNoiseaft = noiseSpec.*abs(Fw).^2;  % ISI power spectrumisiaft = inputSpec.*abs(Fhwout).^2;   % signal power spectrumsignalaft = inputSpec.*abs(Fhwin).^2;    % subband SNRs after equalizationSNRi = signalaft./( colorNoiseaft + isiaft ); %SNRi(MFBi<SNRi) = MFBi(MFBi<SNRi);% calculate geometric SNRs, bit/symbols and bit/second with equalizer[geoSNRfinal bDMTfinal RDMTfinal] = geosnr(SNRi(used),margin,codingGain,N,Nb,fs);% calculate geometric SNRs, bit/symbols and bit/second for upper bound[geoSNRmfb bDMTmfb RDMTmfb] = geosnr(MFBi(used),margin,codingGain,N,Nb,fs);save perf