Research On Mobile Delay Tolerant Network Routing Algorithm Based On Social Attributes

Due to the unique "store-carry-forward" mode of mobile Delay Tolerance Network(DTN),which reduces the dependence of wireless communication technology on basic communication facilities,and can meet the communication requirements under extremely harsh conditions such as intermittent connection,high transmission delay,etc.DTN has a wide range of applications in interstellar networks,wireless sensor networks,emergency disaster relief,etc.Routing technology is the basis of all network related technologies.However,the existing DTN routing algorithm has the following shortcomings:Firstly,the resources of network communication nodes deployed in harsh environments are severely limited,but the existing DTN routing algorithm requires nodes to maintain and process information from other nodes,and has high requirements on the storage of computing resources and energy consumption of nodes.Secondly,the existing DTN routing algorithms lack self-adaptability and resource-aware ability,and can not adapt to the changes of network status.Finally,the decrease of node interaction frequency in the sparse network scenario will increase the difficulty of information collection,which will seriously affect the key performance indicators of existing DTN routing algorithms.In view of the above problems,this paper focuses on the research of mobile delay tolerance network routing algorithm based on social attributes,and its main contributions are as follows:(1)Research on DTN adaptive routing algorithm based on node historical encounter frequencyIn order to solve the problem that node resources are severely limited and the existing DTN routing algorithm cannot adapt to the change of network state,this paper proposes an adaptive DTN routing algorithm based on node relationship tree(AR-RT).Each node constructs and maintains its own relationship tree based on the frequency of historical encounters,and the size of the relationship tree is not affected by the network size,effectively reducing the maintenance burden of nodes.In the forwarding stage,different forwarding strategies are adopted according to the relationship tree,which can improve the forwarding efficiency and reduce the delay,reduce the impact of flooding,reduce the network overhead,and achieve more targeted forwarding.In addition,the algorithm designs an adaptive control mechanism for the generation of message sources.The source node will dynamically control the maximum number of message replicas according to its own cache occupancy,so that the node can make negative feedback to the changes in the network environment,effectively improving the adaptability of the algorithm.Simulation experiments indicate that the AR-RT algorithm proposed in this paper has obvious advantages over the existing DTN routing algorithm in terms of the key indicators such as average delay,average hop count and message delivery rate.(2)Research on DTN routing algorithm under low sampling rate based on cognitive modelAiming at the problem that the convergence time of the algorithm proposed in(1)increases and the message delivery rate also decreases significantly due to the decrease of node interaction frequency in the sparse network scenario,and considering the difficulty in predicting the social relationship between nodes under the condition of low sampling rate and the problem that some nodes reserve their own resources and refuse to relay and forward,this paper proposes a cognitive model-based DTN social routing algorithm(CMSR).Firstly,CMSR algorithm designs a social relationship prediction mechanism by introducing cognitive models in the field of social networks,which can accurately predict the social relationship between all nodes under the condition of low sampling rate.Secondly,nodes are differentiated according to their own characteristics,which are divided into normal nodes and selfish nodes,to reduce the potential risk of message routing due to some nodes dropping messages or not forwarding messages in time.Finally,in the forwarding process,other nodes in the network are classified according to the cognitive matrix according to the social relationship with the destination node,and different message forwarding strategies are adopted for different types of nodes.Simulation experiments indicate that the CMSR algorithm proposed in this chapter can improve the message delivery success rate and reduce the average number of hops while ensuring less network overhead compared with the existing routing algorithms in the network scenario with sparse nodes.