Deep learning decoding method of distributed joint source channel coding system

Abstract

The invention discloses a deep learning decoding method of a distributed joint source channel coding system, and relates to data transmission in electronic communication. Comprising the following steps: 1) determining a training sample set for decoding of the distributed joint source channel coding system; 2) determining a deep learning decoding model; 3) determining a loss function of the deep learning network model; 4) determining the position of a hidden layer needing to be added with an information interaction unit and inserting the hidden layer of the information interaction unit; 5) determining an activation function of a hidden layer of the deep learning decoding training model; 6) determining a training model in a gradient descent mode of the input training sample set, and adjusting training parameters to determine a final network model; and 7) inputting a sample set needing to be decoded into the model to complete decoding of the information source of the distributed joint information source channel system. Parallel decoding can be realized for a distributed joint information source channel coding system, and the correlation between distributed information sources can be mined, so that the decoding performance of the system is improved.

Description

technical field [0001] The invention relates to data transmission in electronic communication, in particular to a deep learning decoding method of a distributed joint source channel coding system based on LDPC codes. Background technique [0002] The fifth-generation 5G cellular network has the characteristics of high speed, low latency and supports the interconnection of a large number of devices. 5G promotes a large number of practical applications of the Internet of Things. At the same time, the emergence of the Internet of Things has also promoted the exponential growth of 5G networks. Among them, the sensor network can realize data transmission with high accuracy and low power consumption, which is a key part of the Internet of Things. The distributed joint source channel coding system is a technology that compresses and encodes the relevant outputs of multiple sensors without communicating with each other and performs joint decoding at the decoding end. It combines ch...

AI Technical Summary

Problems solved by technology

But the decoding network structure for distributed source channel coding system has not been proposed yet

Images