Hard-decoding vector quantization using Bayes estimator for Rayleigh channel

Vector quantization is decoded using a Bayes estimator. The hard-decision Bayes estimator is used to reduce the complexity involving the evaluation of transition probabilities for the soft-decision Bayes estimator. In the range of low to moderate average channel signal-to-noise ratio, the hard-decision Bayes estimator based source decoder can achieve more than 1 dB improvement over the conventional mapping based source decoder. With great reduction in complexity, the hard-decision Bayes estimator can achieve a performance that is just about 0.4 dB below the optimal soft-decision Bayes estimator.