Transactions Papers Effective Informed Dynamic Scheduling for Belief Propagation Decoding of LDPC Codes

Ying Gong,Xingcheng Liu,Guangjie Han
2011-01-01
Abstract:The simultaneous flooding scheduling is popular for Low-Density Parity-Check (LDPC) Belief Propagation (BP) decoding. Non-simultaneous sequential scheduling is superior to the flooding scheduling, and asynchronous dynamic scheduling has better FER performance than the sequential scheduling. However, all strategies encounter the trouble of locating the error variable node. This paper proposes an informed dynamic scheduling strategy, which utilizes the instability of the variable node and the residual of the variable-to-check message to locate the message to be updated first. The informed dynamic scheduling overcomes the trapping sets effectively. This paper also designs an informed dynamic scheduling strategy with adaptivity to pass more messages in parallel, which effectively postpones the influence of cycles in the Tanner graph. In some sense, the strategy lengthens cycles. Simulation results show that the two informed dynamic scheduling strategies outperform other algorithms.
What problem does this paper attempt to address?