认知无线网络频谱共享中的强化学习和深度学习

Reinforcement Learning and Deep Learning in Spectrum Sharing of Cognitive Radio Networks

认知无线电(Cognitive Radio)技术是为了克服无线网络快速发展带来的频谱稀缺问题而发展起来的。认知无线网络(CRNs)通过异构无线架构和频谱共享(Spectrum Sharing)技术为移动用户提供高带宽。同时,由于可用频谱的波动性以及不同网络效用的要求,在认知网络中需要机器学习技术来解决各类问题。机器学习旨在使系统感知和评估可用资源,自主适应感知的无线环境,并重新配置其运行模式,以最大限度地利用可用资源。在这篇文章中,我们重点讨论机器学习技术在频谱共享领域的应用。

Cognitive Radio technology is developed to overcome the shortage of spectrum caused by the rapid development of wireless networks.Cognitive Radio Networks(CRNs)provide high bandwidth for mobile users through heterogeneous wireless architecture and spectrum sharing technology.At the same time,because of the fluctuation of available spectrum and the requirements of different network utility,machine learning technology is needed to solve various problems in cognitive networks.Machine learning aims to enable systems to sense and evaluate available resources,autonomously adapt to a perceived wireless environment,and reconfigure its operating mode to make the most of available resources.This paper focuses on the application of machine learning technology in the field of spectrum sharing.