On fast exhaustive search of the minimum distance of linear block codes

Abstract

Code design and assessment can be a computationally intensive task because of billions operations needed for code evaluation. Therefore, it is important that fast algorithms are devised for minimum distance. We propose a scalable variant of a known algorithm for exhaustive search of the minimum distance. The proposed algorithm is based on smart thread management in order to speed up the exhaustive search. The algorithm is particularly efficient when applied in combination with genetic algorithms for code design as a part of code evaluation and can be easily deployed on grid or cloud computing systems.