Research On Cross-layer Optimization Technology Of Large-scale Ad Hoc Network

With the rapid development of mobile Ad Hoc network technology,the requirements for network performance are getting higher and higher.The continuous increasing of network scale brings challenges to the improvement of network performance.In recent years,cross-layer technology has gradually become an important research direction for experts and scholars because of its high efficiency,reliability,and ability to reduce network information redundancy in solving wireless network problems.How to use cross-layer technology to improve the performance of large-scale Ad Hoc network has also become a research hotspot.The design of protocols has a key influence on the performance improvement of Ad Hoc network.After analyzing the functions of each protocol layer,discussing the method of implementing cross-layer design,and comparing the existing MAC layer protocol and cluster networking protocol,this paper utilizes the idea of cross-layer optimization to optimize the performance of large-scale Ad Hoc network by improving MAC protocol and cluster networking protocol.In order to solve the problems of the traditional MAC protocol,such as low channel utilization,poor distribution fairness,and unsuitable for the network with unbalanced load,this paper proposes an improved FPRP(Five-Phase Reservation Protocol)based on a non-cooperative game model called NG-FPRP(Non-cooperative Game-FPRP).In this protocol,a non-cooperative game model with cross-layer design idea is used to model the slot allocation mechanism.It uses the characteristics of node mobility at the physical layer and the business information of the application layer to carry out cross-layer information fusion to construct a revenue function.The Nash equilibrium solution obtained by the model is used to improve the process of protocol timeslot competition.The NG-FPRP protocol was simulated and verified on the OMNeT++simulation platform.The simulation results show that NG-FPRP can also effectively increase the success rate of nodes in competition,increase the packet delivery rate,and reduce the average delay of the node in the scenario of uneven network load.Aiming at the problems of high network overhead and short life cycle of traditional clustering networking protocols,a cross-layer optimized clustering networking protocol based on gray wolf optimization(Grey Wolf Optimization Clustering protocol,GWOC)is proposed.It uses the gray wolf optimization algorithm to model and analyze the cluster head election problem.It adopts the idea of cross-layer design,and optimizes the application function by using node features such as energy and mobility of the physical layer and business information of the application layer.It improves the problem of cluster head election,optimizes the process of cluster establishment and improves cluster maintenance process.The protocol was simulated and verified on the OMNeT++simulation platform.The simulation results showed that this protocol prolongs the life of the network and increases the packet delivery rate.It is suitable for large scale Ad Hoc network with diversified services and strong mobility.