- Published 2000

Two types of maximum likelihood decoding algorithms for linear block codes have been proposed. The decoding complexity of the first type (for example, Viterbi type algorithms [1], [2]) is almost independent of the signal to noise ratio, and the complexity of the second type (for example, iterative decoding algorithms [3], [4]) decreases considerably as the signal to noise ratio increases. In this paper, a recursive maximum likelihood decoding algorithm of the second type, called “Top-down RMLD [8-10]", is introduced based on the original recursive maximum likelihood decoding algorithm [5-7] of the first type, abbreviated as ”Bottom-up RMLD", by a lazy evaluation strategy.

@inproceedings{Kasami2000ISIT0,
title={ISIT ' 00 Shannon Lecture : Soft - Decision Recursive Decoding for Binary Linear Block Codes *},
author={Tadao Kasami},
year={2000}
}