46 lines
1.6 KiB
Matlab
46 lines
1.6 KiB
Matlab
function [y,csi] = preHESymbolEqualize(x,chanEst,noiseVar,varargin)
|
|
%preHESymbolEqualize Pre-HE fields frequency domain channel equalization
|
|
%
|
|
% [Y,CSI] = preHESymbolEqualize(X,CHANEST,NOISEVAR) performs
|
|
% minimum-mean-square-error (MMSE) frequency domain equalization using
|
|
% the signal input X, the channel estimate, CHANEST, and noise variance
|
|
% NOISEVAR.
|
|
%
|
|
% Y is an estimate of the transmitted frequency domain signal and is of
|
|
% size Nst-by-Nsym, where Nst represents the number of carriers
|
|
% (frequency domain) and Nsym represents the number of symbols (time
|
|
% domain). It is complex when either X or CHANEST is complex, or is real
|
|
% otherwise.
|
|
%
|
|
% CSI is a real matrix of size Nst-by-Nr containing the soft channel
|
|
% state information.
|
|
%
|
|
% X is a real or complex array containing the frequency domain signal to
|
|
% equalize. It is of size Nst-by-Nsym-by-Nr, where Nr represents the
|
|
% number of receive antennas.
|
|
%
|
|
% CHANEST is a real or complex array containing the channel estimates for
|
|
% each carrier. It is of size Nst-by-1-by-Nr.
|
|
%
|
|
% NOISEVAR is a nonnegative scalar representing the noise variance.
|
|
%
|
|
% [Y,CSI] = preHESymbolEqualize(...,NUMSUBCHANNELS) merged demodulated
|
|
% symbols and channel estimates for each subchannel before equalization.
|
|
|
|
% Copyright 2018-2020 The MathWorks, Inc.
|
|
|
|
%#codegen
|
|
|
|
if nargin>3
|
|
numSubchannels = varargin{1};
|
|
else
|
|
numSubchannels = 1;
|
|
end
|
|
|
|
% Treat each 20 MHz subchannel as a receive antennas for diversity
|
|
[symMerge,chanMerge] = wlan.internal.mergeSubchannels(x,chanEst,numSubchannels);
|
|
|
|
% Equalize
|
|
[y,csi] = helperSymbolEqualize(symMerge,chanMerge,noiseVar);
|
|
|
|
end |