基于深度学习的无线物理层关键技术研究综述

A survey of wireless physical layer key technology based on deep learning

目前深度学习在通信系统的上层中得到了广泛应用,随着技术不断发展,深度学习正在向通信系统的底层推进。为解决传统算法计算效率低、复杂度高等问题,深度学习已经被应用到无线通信物理层关键技术中。对深度学习在无线通信物理层关键技术中的应用进行了综述性讨论,包括深度学习定义、深度学习神经网络介绍、基于深度学习的无线物理层关键技术等。分析表明,深度学习与无线通信系统之间存在结合点,在传统的通信模块或算法中加入用深度学习训练的可学习的参数是当前比较具有竞争力的一种设计方案,有必要对此进行深入研究。

Deep learning is now widely used in the upper layers of communication systems.Deep learning is advancing to the bottom of the communication system as technology continues to evolve.Deep learning has been applied in wireless physical layer key technology to solve the problems of low computational efficiency and high complexity of traditional algorithms.The application of deep learning in wireless communication physical layer key technology is discussed in this paper,including the definition of deep learning,the introduction of deep learning neural network and deep learning-based wireless physical layer key technology.The analysis in this paper shows that there are junctions between deep learning and wireless communication systems.It is a relatively competitive design that the learnable parameters of deep learning training are added into traditional communication modules or algorithms,which is necessary to conduct in-depth research.