Polar code decoding method and device based on neural network

Abstract

The embodiment of the invention provides a polar code decoding method and device based on a neural network. The method comprises the following steps: obtaining a polarization code to be decoded, the polarization code being obtained by performing polarization coding on an uncoded sequence, the uncoded sequence being composed of information bits of K bits and frozen bits of (N-K) bits, N = 2n, n being a positive integer, and K being a positive integer less than N; inputting a polarization code to be decoded into the neural network model to obtain a decoding sequence; wherein the neural network model is obtained by training based on a preset training set, and the preset training set comprises a sample polarization code and a sample uncoded sequence. By applying the technical scheme provided by the embodiment of the invention, after the neural network model is trained, the error between the decoding sequence of the sample polar code and the uncoded sequence of the sample is adjusted to bewithin an acceptable range, so that the bit error rate of polar code decoding can be reduced; in addition, parallel decoding can be carried out by using the neural network model, and the delay time ofpolar code decoding is reduced.

Description

technical field [0001] The invention relates to the field of communication technology, in particular to a neural network-based polar code decoding method and device. Background technique [0002] In the field of communication technology, polar codes are a new breakthrough. When the code length is infinite, the binary input BMCs (Binary-input Discrete Memoryless Channel, discrete memoryless channel) can be achieved under the SC (Successive Cancellation, serial cancellation) decoding algorithm. channel) channel capacity. When the code length is limited, SCL (Successive Cancellation List, Serial Cancellation List) decoding assisted by CRC (Cyclic Redundancy Check, Cyclic Redundancy Check Code) can achieve a lower density than turbo code and LDPC (Low Density Parity Check, parity code) codes for better performance. Existing polar code decoding algorithms include SC-based algorithms and BP (Belief Propagation, Belief Propagation)-based algorithms. Among them, the serial decodin...

AI Technical Summary

Problems solved by technology

Existing polar code decoding algorithms include SC-based algorithms and BP (Belief Propagation, Belief Propagation)-based algorithms. Among them, the serial decoding characteristics of SC-based algorithms limit its throughput and have a high decoding delay

Although the BP-based algorithm can be decoded in parallel, the complexity of BP decoding is high, and there is a certain gap between the decoding performance and the SC-based algorithm.

Images