An efficient decoding scheme for improved throughput in three dimensional turbo product codes based on single parity code

This study proposes an effective decoding method that can be used when the existing two dimensional Turbo Product code (TPC) with extended Hamming code is expanded to three dimensional TPC by applying Single Parity code (SPC). In the multidimensional TPC, the SPC is mainly used for increasing the code rate of the axis with short code length. However, the SPC is unable to correct errors, and therefore, its application only is unlikely to improve the performance of the three dimensional TPC. Moreover, its expanded dimensions decrease the throughput. To solve this problem, three types of decoding schemes are suggested. The first scheme is applied to form of mixed iterative decoding scheme that serial and parallel decoding method are used together. Secondly, Normalized Min-Sum (NMS) algorithm is used in axis with shortest code length. Finally, Partial sorting networks reduces clock consumption of least reliable values selection in Chase-Pyndiah algorithm.