关键词: 无线网络, 保密率, 中继选择, 解码转发, 深度学习

Abstract: Under the background of the integration of information and intelligence, multihop wireless communication network, as a key carrier connecting multiuser communication equipment, is often faced with the threat of information eavesdropping in the process of longdistance wireless relay transmission. In order to improve the information security in the wireless communication process, this paper proposes a high security information transmission method based on deep learning, whose core is to ensure the security of the information transmission process through optimal relay selection. Firstly, the artificial neural network model is established, and the channel state data in various communication environments are used to train the model, so as to select the multihop relay scheme and obtain the maximum security rate under the decode and forward (DF) relay constraint. Simulation results show that compared with the traditional exhaustive search method, the proposed method can achieve a security performance close to 0.2b(Hz·s-1). Moreover, since the model training is preliminary work, the results will be returned directly according to the channel information in practical application, so the calculation time can be significantly reduced.

Key words: wireless network, security rate, relay selection, decodingandforwarding, deep learning