In order to solve the problem that the existing Polar code decoding algorithm based on neural network can only decode short codewords (codewords length N≤64), a new decoding algorithm using convolution neural network for long codewords (N≥512) was put forward. Instead of using a fixed drawn from the data sample proportion as training set and test set, the neural network parameters epoch and batch were used to control the neural network input, and the data explosion problem caused by the long codewords was solved. In addition, the influence of batch and epoch parameters on the decoding performance of convolution neural network was explored and the performance difference of neural network using different activation functions were investigated. Simulation results show that with the traditional SCL (successive cancellation list, L=2) decoding algorithm, convolution neural network in low signal-to-noise ratio on the performance is better than that of SCL (L=2), in high signal-to-noise ratio and SCL (L=2) algorithm of similar performance, and the larger the training data set, the better the decoding performance of neural network.
In order to solve the problem that the existing Polar code decoding algorithm based on neural network can only decode short codewords (codewords length N≤64), a new decoding algorithm using convolution neural network for long codewords (N≥512) was put forward. Instead of using a fixed drawn from the data sample proportion as training set and test set, the neural network parameters epoch and batch were used to control the neural network input, and the data explosion problem caused by the long codewords was solved. In addition, the influence of batch and epoch parameters on the decoding performance of convolution neural network was explored and the performance difference of neural network using different activation functions were investigated. Simulation results show that with the traditional SCL (successive cancellation list, L=2) decoding algorithm, convolution neural network in low signal-to-noise ratio on the performance is better than that of SCL (L=2), in high signal-to-noise ratio and SCL (L=2) algorithm of similar performance, and the larger the training data set, the better the decoding performance of neural network.
In order to overcome the problems of high bit error rate, delay and complexity, research on CNN polar decoding algorithm is proposed. This method uses CNN as the decoder, analyzes the difference between the traditional algorithm and the CNN decoder, adjusts the network parameters, obtains the best parameters, makes the decoder optimal, and in the channel coding scheme, the encoding and decoding complexity of polarized codes is low, and it has been strictly proved that they can reach the Shannon limit. However, there are a lot of shortcomings in the existing traditional decoding algorithms, With the maturity of deep learning, the application of deep learning to the field of communication, deep learning has advantage of powerful computing, and neural network after training is static, once only need data through the network, so the deep learning was applied to the polarization code decoding process, can effectively reduce the decoding delay and improve the efficiency of decoding, this paper mainly studies convolution decoder of neural networks in all aspects of performance, can get the conclusion by experiment, with the increase of the number of iterations network decoding ability improved, error performance is getting better and better, and CNN has obvious advantages.
Known for their capacity-achieving abilities, polar codes have been selected as the control channel coding scheme for 5G communications. To satisfy high throughput and low latency, belief propagation (BP) is ideal as the decoding algorithm due to its nature of parallel processing. However, the error performance of BP is in general worse than that of enhanced successive cancellation (SC). Recently, bit-flipping (BF) mechanism is applied to BP decoding to lower the error rate. However, its trial-and-error process results in longer latency. In this work, we propose a convolutional neural network-aided bit-flipping (CNN-BF) mechanism to further enhance BP decoding. With carefully designed input data and model architecture, our proposed CNN-BF can achieve better error correction capability with less flipping attempts than prior works. It also achieves a lower block error rate (BLER) than SC list (SCL).
Known for their capacity-achieving abilities, polar codes have been selected as the control channel coding scheme for 5G communications. To satisfy the needs of high throughput and low latency, belief propagation (BP) is chosen as the decoding algorithm. However, in general, the error performance of BP is worse than that of enhanced successive cancellation (SC). Recently, critical-set bit-flipping (CS-BF) is applied to BP decoding to lower the error rate. However, its trial and error process result in even longer latency. In this work, we propose a convolutional neural network-assisted bit-flipping (CNN-BF) mechanism to further enhance BP decoding of polar codes. With carefully designed input data and model architecture, our proposed CNN-BF can achieve much higher prediction accuracy and better error correction capability than CS-BF but with only half latency. It also achieves a lower block error rate (BLER) than SC list (CA-SCL).
By adopting successive cancellation list decoding (SCL), polar codes demonstrate competitive error correction performance over LDPC and Turbo codes. However, SCL decoding suffers from high computational complexity and long decoding latency, especially when the list size is very large. Successive cancellation flip (SCF), as another decoding algorithm that can achieve high error correction performance, has a complexity that is close to that of successive cancellation (SC) decoding. With the observation that SCL and SCF decoding are similar at giving more chances to inspect possible codewords simultaneously or sequentially, a novel hybrid decoder is proposed in this paper, which essentially combines the ideas of SCF and SCL decoders. Moreover, in order to compensate for the degradation of performance caused by the reduction of path splitting and further reduce the decoding latency, the convolutional polar codes are adopted with a designed bit-flipping set. Simulation results demonstrate that the proposed decoder achieves the reduction of decoding latency while attaining better performance than conventional CRC-aided SCL decoder.
In order to improve transmission reliability and flexible cooperation in optical communication, rate-compatible punctured Polar codes have become a research hotspot. Aiming at the problem that the traditional decoding performance and transmission efficiency is limited, based on deep learning, a rate-compatible punctured Polar code decoding scheme is studied. We use convolutional neural network model as the basic structure of rate-compatible Polar code decoder. The log likelihood ratio values of the received sequence are input into the decoder for training. Simulation results show that the proposed decoder outperforms the traditional punctured Polar code decoder under high signal-to-noise ratio.
Deep learning recently shows strong competitiveness to improve polar code decoding. However, suffering from prohibitive training and computation complexity, the conventional deep neural network (DNN) is only possible for very short code length. In this paper, the main problems of deep learning in decoding are well solved. We first present the multiple scaled belief propagation (BP) algorithm, aiming at obtaining faster convergence and better performance. Based on this, deep neural network decoder (NND) with low complexity and latency, is proposed for any code length. The training only requires a small set of zero codewords. Besides, its computation complexity is close to the original BP. Experiment results show that the proposed (64,32) NND with 5 iterations achieves even lower bit error rate (BER) than the 30-iteration conventional BP and (512, 256) NND also outperforms conventional BP decoder with same iterations. The hardware architecture of basic computation block is given and folding technique is also considered, saving about 50% hardware cost.