MIMO equalization of 16-QAM signal blocks using an FFT-based alphabet-matched CMA

In this work we investigate the blind equalization of signal blocks of data, which have been transmitted over a multiple-input/multiple-output channel using 16-QAM modulation. We apply the Fast Fourier Transform (FFT) in order to improve the efficiency of the blind equalizer update. Our equalizer is based on a cost function, which is specified as the sum of the constant modulus algorithm (CMA) and the alphabet-matched algorithm (AMA). The equalization is done using a block processing multistage approach as was introduced in [3], but with an FFT implementation. In this work, a 3-input/7-output MIMO communication system is considered, and the performance of the MIMO equalizer is analyzed using Monte Carlo simulations. Performance is measured based on the average symbol error rate (SER) and the total number of equalized cases. In addition, initial comparative results are provided for a fixed number of transmit antennas, but with varying numbers of receive antennas.