Sequential decoding of convolutional codes in channels with intersymbol interference

Sequential sequence estimation for communications channels with convolutionally coded input sequences and intersymbol interference (ISI) is proposed. The receiver consists of a whitened matched filter (WMF) and a vector sequential decoder. The metric of the vector sequential algorithm (SA) is derived. Analytical results for error probability and computational complexity of the sequential decoder are presented. The error probability is found to be dominated by the minimum Euclidean distance between the possible output sequences and the computational distribution is Pareto, typical of a sequential decoder. The computational cut-off code rate R/sub comp/ and its lower bound R/sub comp/' are also derived. Numerical simulation results of the error probability and the computational complexity are presented. Comparing with the Viterbi algorithm (VA) reveals that the SA essentially has the same error performance as that of the VA while its average computational complexity is one to two orders of magnitude less than that of the VA.<
>