Multi-CDS Scheduling Mechanism Based Topology Control Algorithm For Fanet

FANET(Flying Ad Hoc Network)based on unmanned aerial vehicle systems has been a research hotspot in the field of self-organizing networks in recent years.Compared with the ground MANET network,FANET network system has higher flexibility and has a good application prospect in military,rescue and transportation.Because nodes in FANET network have high moving speeds and high dynamic topology changes,link failures are frequent and the robustness of the network is poor,which poses a severe challenge to topology control.Therefore,effective topology control is a key issue in current ad hoc networks.Topology control in FANET mainly solves the problems of node connectivity and routing maintenance.By changing node power to balance the distribution of neighbor nodes,or constructing node attributes through network layered control,the network invulnerability is guaranteed and the transmission efficiency is improved.The common algorithms are clustered and construct a connected dominating set.This paper mainly considers the FANET network connectivity and the resulting routing forwarding redundancy problem.By designing an effective topology control algorithm,it can maintain the normal communication of nodes in the network and reduce the routing forwarding overhead.Constructing the virtual backbone network is an effective method to solve message forwarding and reduce routing overhead.Connected dominating set(CDS)is widely used in self-organizing networks.The existing CDS-based virtual backbone network algorithm can not be directly applied to FANET,mainly in the following three points: The mobile model used in the algorithm,the three-dimensional movement characteristics of FANET nodes.cluster collaboration characteristics do not match;the algorithm overhead is too large;when the topology changes the backbone network maintenance algorithm is single.In order to solve the above problems,considering that the node mobility model has Markov characteristics,this paper proposes a link prediction algorithm based on Markov chain to detect the link connectivity of the backbone network subgraph.By predicting node connectivity,the backbone network node or link failure is determined.Constructing a smaller-scale CDS in the network can reduce the overhead of constructing and maintaining the backbone network and reduce the message forwarding times.The minimum connected span dominating set construction algorithm based on minimum spanning tree can construct the connected dominating set with smaller weights and smaller scales.Therefore,this paper designs an algorithm based on minimum spanning tree for CDS construction.Taking into account the high dynamics and link instability of FANET,a single CDS can ensure network connectivity after the failure of a dominant point or link.Therefore,in order to guarantee the success rate of the maintenance algorithm and reduce the update process overhead,this paper proposes a maintenance algorithm based on multi-CDS scheduling.Based on the link prediction and periodic update of the topology,determine whether the current CDS is valid.When the topology detects or predicts the failure of the backbone network,the backup CDS is selected using the principle of a minimum number of updated nodes.Among them,the link prediction algorithm improves the success rate of the maintenance algorithm,reduces the delay of the maintenance process,and ensures the stability of the network communication.By comparing the typical backbone network maintenance algorithms based on backup nodes,the simulation results show that the Multi-CDS scheduling algorithm has a smaller update overhead and a higher maintenance success rate.