The flip-successive cancellation (SCF) decoding algorithm is a decoding scheme to improve the performance of the SC decoding algorithm under short code length by flipping erroneous bits in initial SC decoding. The degraded performance of the SCF decoding algorithm is usually caused by the wrong locating of the first erroneous bit or additional erroneous bits. To address this issue, we propose a machine learning based multi-flips SC decoding scheme (ML-MSCF), which can improve the performance of the SCF decoding algorithm with multiple flips based on the long short-term memory (LSTM) network and reinforcement learning (RL). Specifically, we use a LSTM network to locate the first erroneous bit when initial SC decoding fails, then the outputs of the LSTM network are used as the action space of RL to identify additional erroneous bits in the followed procedure. Simulation results show that the proposed scheme can achieve performance improvement of 0.2–0.3dB over the stateof-art SCF decoding algorithm on both the bit error ratio (BER) and the frame error rate (FER) with less decoding latency.
A Machine Learning Based Multi-flips Successive Cancellation Decoding Scheme of Polar Codes
2020-05-01
175771 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Adjustable Soft List Decoding for Polar Codes
IEEE | 2019
|A Novel Interleaving Scheme for Polar Codes
IEEE | 2016
|