A kind of ldpc decoding system and decoding method based on semi-supervised deep learning network

Abstract

The invention provides a LDPC decoding system based on a semi-supervised deep learning network, comprising: an input unit for constructing LDPC codewords into a sample set suitable for deep learning network; an unsupervised learning mode, which includes a noise reduction unit for reducing noise of the sample set and a feature extraction and feature mapping unit for carrying out feature extractionand feature mapping on noise-reduced data; and a supervised learning unit that is used for supervised training for input after feature mapping and returns a batch of LDPC codeword information bits tocomplete LDPC batch decoding. The input unit, the noise reduction unit, the feature extraction and feature mapping unit, and the supervised learning unit are connected in turn. The invention also provides a LDPC decoding method based on a semi-supervised deep learning network. According to the invention, the high throughput decoding without iteration realized by using a trained deep learning network can be used to solve the problems of high decoding complexity, delay caused by multiple iterations and low throughput of LDPC soft decision decoding algorithm in the prior art.

Description

technical field [0001] The present invention relates to the technical field of electronic communication, and more specifically, relates to an LDPC decoding system and a decoding method based on a semi-supervised deep learning network. Background technique [0002] LDPC code (Low-Density Parity-Check), that is, low-density parity-check code, is a channel code that has been proven that its error performance can approach the Shannon limit, and its decoding complexity is relatively low and its performance is good. LDPC codes can meet the requirements of mobile communication technology for high data processing speed, high data transmission speed, large-capacity transmission and high-quality transmission. One of the mainstream encoding methods. [0003] Deep learning has proven its powerful recognition, classification and fitting capabilities in speech, image, natural language processing and other applications. Deep learning networks include supervised learning networks, unsuper...

AI Technical Summary

Problems solved by technology

[0005] The object of the present invention is to overcome the shortcoming and deficiency in the prior art, provide a kind of LDPC decoding system and decoding method based on semi-supervised deep learning network, this system and method utilize the well-trained deep learning network to realize the high-speed decoding without iteration. Throughput rate decoding is used to solve the problems of high decoding complexity, delay caused by multiple iterations and low throughput of the existing LDPC soft-decision decoding algorithm

Images