Improved Decoding of linear Block Codes using compact Genetic Algorithms with larger tournament size

Soft-decision decoding is a very important NP-hard problem for developers of communication systems. In this work we propose two new dual domain soft decision decoders that use compact Genetic Algorithm (cGA) with larger tournament size: the first algorithm investigates tournament selection with larger size using mutation, and the second employs higher selection pressure with randomly generated individuals. The obtained results are compared to known previous works and show the effectiveness of using larger tournament size in dual domain soft decision decoding problem. Behind performances analysis, a complexity study is done which shows that both proposed decoders are not very complex in comparison with the standard compact Genetic Algorithm based decoder (cGAD).