Two-Staged Informed Dynamic Scheduling for Sequential Belief Propagation Decoding of LDPC Codes

Abstract

Recent studies have shown that sequential belief propagation decoding of low-density parity-check codes can increase the decoding convergence speed while simultaneously improving the asymptotic performance compared to the conventional flooding scheme. Two of the practical sequential decoding schemes known are the ones by Casado et al. [1] in which informed dynamic scheduling is used for scheduling the sequential updates of the messages. In this letter, we propose a two-staged informed dynamic scheduling that unifies and outperforms the two schemes of [1].