基于Q-learning的动态频谱接入算法研究  

Dynamic Spectrum Access Algorithm Based on Q-learning

作　　者：张亚洲[1] 周又玲[1] Zhang Yazhou;Zhou Youling(College of Information Science and Technology, Hainan University, Haikou 570228, Chin)

机构地区：[1]海南大学信息科学技术学院,海南海口570228

出　　处：《海南大学学报（自然科学版）》2018年第1期9-15,共7页Natural Science Journal of Hainan University

基　　金：国家自然科学基金(61340027);海南省自然科学基金(614222)

摘　　要：设计了一种基于Q-learning的动态频谱接入算法.该算法的利用Q-learning自适应能力强的优势,帮助认知用户感知合适的信道并接入,认知用户通过不断地与环境交互学习选择回报率最高的信道,实现频谱资源的二次利用.最后对该算法进行了仿真,仿真结果表明所提算法能够提高认知用户吞吐量,降低与主用户冲突概率,且能以较快的速度收敛,这对于解决频谱资源利用问题有着重大的意义.In the report, a dynamic spectrum access technology based on Q-learning was proposed. The algorithm takes advantage of Q-learning adaptive ability, and which can help secondary user to perceive the appropriate channels and access. By interacting with the environment, secondary users select the highest rate of return, and which achieve the secondary use of spectrum resources. The simulation results showed that the proposed algo- rithm can improve the secondary user＇ s throughput and reduce the probability of collision with the primary user, and the converging speed is also accelerated. The algorithm will be helpful for solving the problem of spectrum resources use.

关 键 词：动态频谱接入 Q-LEARNING 信道选择 吞吐量 冲突概率 

分 类 号：TN99[电子电信—信号与信息处理]