Polar codes have been adopted as the control channel coding scheme in the fifth generation new radio (5G NR) standard due to its capacity-achievable property. Traditional polar decoding algorithms such as successive cancellation (SC) suffer from high latency problem because of their sequential decoding nature. Neural network decoder (NND) has been proved to be a candidate for polar decoder since it is capable of one-shot decoding and parallel computing. In his paper, we propose a residual neural network decoder (RNND) for polar codes. Different from previous works which directly use neural network for decoding symbols received from the channel, the proposed RNND introduces a denoising module based on residual learning before NND. The proposed residual learning denoiser is able to remove remarkable amount of noise from received signals. Numerical results show that on one hand our proposed RNND outperforms traditional NND with regard to the BER performance under comparable latency. On the other hand, although the proposed scheme is a little inferior to the traditional successive cancellation list decoder when the BER performance is under consideration, it reduces significant decoding latency with the aid of the parallel structure of neural network and modern graphical processing units.
Learning to Denoise and Decode: A Novel Residual Neural Network Decoder for Polar Codes
01.11.2020
820200 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
A deep learning approach to denoise luminescence images of solar cells
TIBKAT | 2022
|Study and Design Comprehensive Denoise System for Railway
Tema Archiv | 2013
|Pipelined BP Polar Decoder with a Novel Updated Scheme
Springer Verlag | 2021
|A Novel Interleaving Scheme for Polar Codes
IEEE | 2016
|PARTIAL DECODE-AND-FORWARD OF LDPC CODES FOR ON-BOARD PROCESSING SATELLITE PLATFORM
British Library Conference Proceedings | 2013
|