
In the 5G communication systems, polar code has been adapted as the control channel coding solution in the enhanced mobile broadband (eMBB) scenario. Although different decoding schemes, including belief propagation (BP) and successive cancellation (SC) based algorithms, have been proposed, the decoding complexity as well as the latency are still significant. To address this critical issue, several low-complexity schemes, e.g., the use of simplified decoding operation and stop the decoding operation in earlier stage, have been proposed recently. However, conventional early stopping strategies have to check a pre-defined metric in each iteration, and the associated decoding delay is significant. In this paper, to address this challenge, we proposed a low-complexity BP based decoding scheme, which contains the decodability detection stage and the early stopping prediction stage. The decodability detection stage can identify the codewords in the deep channel fading environment and eliminate the unnecessary decoding operations to reduce the decoding complexity, while the early stopping prediction stage can directly predict the required number of iterations rather than checking the metric in each iteration to avoid the associated decoding delay. Through the above two approaches, our proposed scheme is shown to achieve 71% decoding delay reduction and maintain the same decoding performance as traditional BP, G-matrix, MinLLR schemes.
early stop prediction, Polar codes, BP decoding, deep learning, Electrical engineering. Electronics. Nuclear engineering, decodability detection, TK1-9971
early stop prediction, Polar codes, BP decoding, deep learning, Electrical engineering. Electronics. Nuclear engineering, decodability detection, TK1-9971
| selected citations These citations are derived from selected sources. This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically). | 13 | |
| popularity This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network. | Top 10% | |
| influence This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically). | Top 10% | |
| impulse This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network. | Top 10% |
