Title:Dynamic Layered Decoding Scheduling for LDPC Codes Aided by Check Node Error Probabilities

Abstract:In this study, a new scheduling strategies for low-density parity-check (LDPC) codes under layered belief propagation (LBP) is designed. Based on the criteria of prioritizing the update of check nodes with lower error probabilities, we propose two dynamic scheduling methods: dynamic error belief propagation (Dyn-EBP) and dynamic penalty error belief propagation (Dyn-PEBP). In Dyn-EBP, each check node is restricted from being updated the same number of times, whereas Dyn-PEBP removes this restriction and instead introduces a penalty term to balance the number of updates. Simulation results show that, for 5G new radio (NR) LDPC codes, our proposed scheduling methods can outperform existing dynamic and offline scheduling strategies under various blocklengths and code rates. This demonstrates that prioritizing the update of check nodes with lower error probabilities can lead to higher decoding efficiency and validates the effectiveness of our algorithms.