An Enhanced Chase-Pyndiah Algorithm for Turbo Product Codes

Turbo Product Code (TPC), as a kind of linear block codes with high coding efficiency and great error correction performance, is widely used in wireless communication. To improve the decoding performance of TPC, this paper proposes a new decoding algorithm, which designs a pre-decoding operation before applying the conventional decoding algorithm, i.e., Chase-Pyndiah algorithm. The proposed algorithm exploits the syndrome to select some bits with high confidence in TPC matrix, and alleviates the incorrect decoding of the other unreliable rows or columns by enhancing their soft information. Then, the enhanced soft information matrix is used to effectively improve the decoding performance during the iterations of Chase-Pyndiah algorithm. The simulation results verify the performance advantage of our proposed algorithm when compared with the benchmark.