Semi-Blind Channel Estimation For Linearly Precoded MIMO-CPSC

In this paper, linear preceding is introduced into MIMO-CPSC system to enhance channel estimation. Thanks to the preceding, the input symbols have well-defined statistical correlations. This property is then used for blind channel estimation. By modifying and eigen-decomposing the statistical auto-correlation matrix of the frequency domain received signals, the channels are estimated subject to a matrix (with size K x K, where K is the number of transmitters) ambiguity. Several linear precodings are presented. The precodings do not change the power of a precoded block. Also, there are fast algorithms to implement the precodings and decodings. Simulations show that the method is effective and robust.