Turbo space-time equalization of TCM with receiver diversity .II. Maximum-likelihood detection

For pt.I see ibid., p.552-56 (2000). This paper presents a turbo equalization method for complex TCM signals over frequency selective, multipath fading channels based on receiver antenna array measurements. The channel observed at each array element is described as an equivalent convolutional encoder acting on the interleaved TCM symbols. The received vector signal can be viewed as produced by a serial concatenated encoder and is decoded by an iterative equalizer that employs M-ary soft output Viterbi algorithm (SOVA) as the decoding rule. Since the computational complexity of the equalizer increases with the number of ISI symbols and antennas used in the receiver, an alternative receiver is also considered where the array outputs are first combined through a beamformer and then sent to the equalizer. Both receiver structures are simulated for two dimensional TCM signals such as 8-16 PSK and 16-QAM and the results indicate an improved performance of the diversity receiver.