一种基于动态帧时隙的强化学习频谱切换方法

A dynamic frame time-slot spectrum handoff scheme based on reinforcement learning

频谱切换作为认知无线电中的重要一环,在多二级用户(Secondary Users, SUs)存在的情况下,频谱切换的方法设计一直是研究的重心。本文在认知无线电网络中存在多个主用户(Primary Users, PUs)和二级用户的情况下,设计了一种基于动态分配帧时隙(Frame Time-slot, FTS)的强化学习频谱切换方法。在本文方法中,我们针对认知无线电Ad Hoc网络中的二级用户,根据单位时间内二级用户的碰撞次数,来动态调整该二级用户的帧时隙长度,从而缩短二级用户传输所用时长,降低其丢包率。仿真结果表明,本文设计的方法与基于固定帧时隙的强化学习频谱切换方法相比,二级用户平均传输时长减少13.1%,整体丢包率减少11.1%。

As an important aspect of cognitive radio technology, spectrum handoff scheme design has always been the focus of research in the presence of multiple secondary users (SUs). In this paper, under the scenario of multiple primary users (PUs) and secondary users in cognitive radio networks, a dynamic frame time-slot (FTS) based on reinforcement learning spectrum handoff scheme is designed. In the proposed spectrum handoff scheme, we dynamically adjust the frame size of the secondary user in cognitive radio Ad Hoc network according to the collision times of the secondary user in unit time, so as to shorten the SUs’ transmission time and reduce the packet loss rate. Simulation results show that compared with the fix frame time-slot reinforcement learning spectrum handoff scheme, the scheme we proposed can reduce the average transmission time of secondary users by 13.1%, and the overall packet loss rate by 11.1%.