Reliability-Based List Decoding with Multiple Biases

摘要

In this work, a new approach to reliability-based list decoding is introduced and analyzed. This reliability-based decoding (RBD) algorithm repeatedly uses biased reliability values to construct a list of candidate codewords. As a result, this method makes use of multiple information sets in a stochastic way. Compared to existing approaches, it provides a more efficient RBD algorithm to achieve near maximum likelihood decoding with a proper choice of the bias value. It can be combined with any RBD algorithm and in each case, a tight performance analysis can be derived.