Decode to channel binary block codes based on neural networks and genetic algorithm

In this article, an error-correction decoding technique using Hopfield neural networks (HNN) and genetic algorithm (GA) is presented. First, based on the elementary relationship between HNN and binary block codes, a new effective algorithm is put forward that can perform maximum likelihood decoding to all cyclic block codes (for example, (23,12) Golay code sets) by use of high-order mutual connecting neural networks. The decoding to each received code word is processed in parallel, and thus it is suitable for high-rate data communication. Second, this HNN decoding algorithm is combined with the genetic algorithm and a scheme decoding general binary block code is developed. The main merit of the proposed HNN and GA decoding technique is that it can efficiently escape the local minima of the energy function and completely achieve correct decoding for all binary block codes.