To solve the problem that the preset single media access control(MAC)protocol is difficult to meet the diversified service requirements of FANET.an adaptive MAC protocol based on Q Learning for FANET(FQL-AMAC)was proposed.The joint control of the two benchmark protocols was used to automatically select and switch to the MAC protocol with significant quality of service(QoS)according to the current network conditions.A single network performance index was optimized to achieve the local optimum of QoS,and the entropy method was used to fuse the throughput and delay,and the comprehensive performance index was proposed to construct the reward function,so as to approach the global optimum of QoS.The results show that FQL-AMAC can effectively select the best protocol to run,and the performances of throughput,delay and compre-hensive performance are better than that of the existing protocols.
关键词
飞行自组织网络/媒体接入控制/多样化业务/自适应选择/Q/Learning/熵值法/综合性能
Key words
flying Ad Hoc network/media access control/diversified service/adaptive selection/Q Learning/entropy method/comprehensive performance
维普
万方数据