Optimizing Mimo Equalization Of Qam Signals Based On Constant Modulus Algorithm And Alphabet Matched Algorithm

Abstract

We consider our implementation of a blind adaptive CMA+AMA equalizer, which is being used to equalize blocks of 16-QAM signals transmitted over a 3- input/7-output MIMO channel. In this work, the MIMO signal equalization is based on a cost function, which is defined as the sum of the constant modulus algorithm (CMA) and alphabet-matched algorithm (AMA). We have modified the Toeplitz matrix of the received signal used to determine the output of the equalizer in our simulator, and we do a comparative analysis of the performance of this MIMO system based on the total number of equalized cases and average symbol error rates (SER). The MIMO channel output signals are equalized by a multistage process, with number of stages equal to the total number of transmit antennas. At each stage of equalization, the captured source is cancelled from the received signal block, and an estimate of the associated path in the MIMO channel is computed. The original implementation of this MIMO simulator is being updated to better handle the end-effects of the convolution process, and this affects the computation of the channel estimate and the entire equalization process.