The traditional Successive Cancellation(SC)decoding algorithm for large kernel matrix polar codes has high computa-tional complexity.This paper uses a trellis to reduce the complexity of the SC decoding algorithm for large kernel matrix polar codes.Firstly,this paper establishes the connection between the internal operations of the SC decoding algorithm and the trellis.Then,it estab-lishes the corresponding trellis-based replacements for internal operations.Finally,based on the BCJR(Bahl,Cocke,Jelinek,Raviv con-struction)trellis,the minimum trellis for internal operations of the SC kernel is constructed.This approach effectively reduces the com-putational load of the algorithm.Simulation results show that for 3×3 kernel polar codes with a length of 243 and a code rate of 1/2,this method reduces the running time by 79.14％compared to direct calculation,savings of 14.2％in computing costs.
维普
万方数据