A STUDY OF VECTOR QUANTIZATION FOR NOISY CHANNELS

Several issues related to vector quantization for noisy channels are addressed. An algorithm based on simulated annealing is developed for assigning binary codewords to the vector quantizer codevectors. It is shown that this algorithm could result in dramatic performance improvements as compared to randomly selected codewords. Also, a modification of the simulated annealing algorithm for binary codeword assignment is developed for the case where the bits in the codeword are subjected to unequal error probabilities (resulting from unequal levels of error protection). An algorithm for the design of an optimal vector quantizer for a noisy channel is briefly discussed and its robustness under channel mismatch conditions is studied. Numerical results for a stationary first-order Gauss—Markov source and a binary symmetric channel are provided. It is concluded that the channel-optimized vector quantizer design algorithm, if used carefully, can result in a fairly robust system with no additional delay. Finally, the case in which the communication channel is nonstationary (as in mobile radio channels) is studied and some preliminary ideas for quantizer design are presented. © 1990 IEEE