Parallel optimization design of LDPC decoder with variable normalization factor

In order to solve the problem of poor decoding performance of the fixed normalization factor in the normalized minimum sum algorithm, the article proposes a variable normalization factor LDPC decoder, which improves the decoding performance compared to the fixed normalization factor. 0.2dB. At the same time, in order to solve the problems of low throughput and high storage resource usage of some parallel LDPC decoders, the parallelism of the system is increased by adopting the intra-layer parallel structure, and the consumption of storage resources is saved by time-division multiplexing of node data cache units. Experimental results show that compared with the traditional partially parallel structure, the decoder designed in this paper uses 13% more LUT resources, reduces storage resource consumption by 38%, and increases the decoding throughput by 14.5M.