Research On Self-adaptive MAC Protocols For UAV Ad Hoc Networks

With the development of wireless communications and Unmanned Aerial Vehicle(UAV)technologies,UAV communications have attracted a great deal of interest among scholars at home and abroad.UAV Ad Hoc Networks(UAVNET)have become a research hotspot due to its characteristics of high dynamic,rapid deployment,scalability and broad application prospects in battlefield reconnaissance,emergency relief,forest monitoring and other aspects.UAVNET MAC protocol plays an important role in node access,channel utilization and network delay.At the same time,the high dynamic characteristics of UAV and the rapid change of network topology bring new challenges to the design of MAC protocol.Therefore,it is of paramount importance to design self-adaptive MAC protocol considering the characteristics of UAV.In this thesis,the self-adaptive MAC protocols based on priority grouping and intelligent backoff mechanism are proposed respectively,and the performances of the proposed protocols are analyzed by simulation.The main works of this thesis are as follows.1)A self-adaptive MAC protocol based on priority grouping is proposed to meet the requirements of quality of service for different traffic in UAVNET.Different service packets are grouped into different priority queues,different access probabilities are set,priority and frame length of packets are considered comprehensively,and the size of node contention window is dynamically adjusted,so as to realize adaptive access control.The access process is modeled theoretically by Markov chain,and the access success probability of node is deduced.Simulation results show that the proposed protocol has higher access success probability than the traditional MAC protocol.2)A self-adaptive MAC protocol based on intelligent backoff mechanism is proposed.Firstly,Markov decision process is used to model the backoff process of UAV node.Considering the number of backoff times and the size of the current packet,the adjustment strategy of the contention window is determined,and the adjustment process of the competition window is modified and updated based on Actor-Critic mechanism,so as to realize the intelligent self-adaptive backoff process.The simulation results show that the proposed protocol has good self-adaptive performance in the scenarios with large number of nodes and long data frames,and can effectively improve the performance of channel utilization,node fairness and average backoff delay.