Minimum Symbol-Error Rate Equalization Based Turbo Receiver for Underwater Acoustic Communications

The minimum-symbol-error-rate (MSER) channel equalizer suffers from slow convergence when the channel has long delay spread. To solve this problem, in the paper we present an iterative turbo receiver by combining the linear MSER equalizer and the MAP channel decoder. With iterative information exchanges between the equalizer and the channel decoder, the turbo receiver updates the equalizer in each iteration. Simulations based on underwater acoustic channels with more than 140 taps show that the proposed scheme converges within a few turbo iterations, and outperforms the conventional minimum-mean-square-error (MMSE) based equalizer.