A Study on Quantization for BP Decoding of Polar Codes

Jian Shu Gao, Kai Niu, Chao Dong

A deep learning method for improving the performance of polar belief propagation (BP) decoder equipped with a one-bit quantizer is proposed. The method generalizes the standard polar BP algorithm by assigning weights to the layers of the unfolded factor graph. These weights can be learned autonomously using deep learning techniques. We prove that the improved polar BP decoder has a symmetric structure, so that the weights can be trained by an all-zero codeword rather than an exponential number of codewords. In order to accelerate the training convergence, a layer-based weight assignment scheme is designed, which decreases the amount of trainable weights. Simulation results show that the improved polar BP decoder with a one-bit quantizer outperforms the standard polar BP decoder with a 2-bit quantizer and achieves faster convergence.

Chen Yi-Ru, Huang–Chang Lee, Lin Mao-Chao

Ziad Ibrahim, Yasmine A. H. Fahmy

Several researchers are interested in polar codes for the sake of their capacity which nearly catches memoryless channels capacity. They have been used as part of the 5G technology. Polar decoders like belief propagation decoders showed unfavorable performance during limited iterations. Nowadays, Machine-learning (ML) techniques are used to enhance the efficiency of polar codes decoders. Recurrent neural network belief propagation (RNN-BP) decoders proved better efficiency in a lower number of cycles. In this paper, a performance scheduling learning technique is proposed. This technique achieves a coding gain of 0.2 dB for the RNN-BP decoder at BER of 2*10−4. A new approach is also introduced in the quantization step. In this approach, the quantization is done after the learning process is finished instead of after each iteration. This approach fixes the issue of model convergence. These combined approaches show up to 0.45 dB improvement at BER of 2*10−4.

Dongsheng Wu, Qingshuang Zhang, Yingxian Zhang

Gou Hosoya, Hiroyuki Yashima

Xinyi Gu, Mohammad Rowshan, Yixuan Xie +1 more

Belief propagation (BP) decoding of polar codes benefits from the high parallelism in hardware implementation and consequently provides high throughput compared to successive cancellation-based decoding algorithms. As the evolved log-likelihood ratios (LLRs) in the decoding process have a wide dynamic range, the quantization error can degrade the error correction performance of the BP decoder. To reduce the quantization error in a uniform quantization scheme, we may need to use more bits for each LLR value which is not desirable due to the required memory space. In this paper, we design an efficient hardware architecture for non-uniformly quantized LLR messages in BP decoding where the arithmetic operations on logarithmically compressed messages are replaced with the mapping to precomputed results in lookup tables. By employing 5-bit non-uniform quantization, the designed BP decoder architecture reduces the required memory space by 37.5 percent compared to 8-bit uniform quantization while additionally improving the block error rate (BLER) by more than 0.1 dB in high SNR regimes. Compared to uniform quantization, the BLER improvement of the designed architecture is up to 0.4 dB when both quantization schemes use 5 bits for LLRs.