Research On Probability Routing Algorithm And Message Confirmation Mechanism Based On Utility Value In Opportunity Networks

Opportunity network is a new type of network which is different from the traditional network.In this kind of network,the communication between nodes presents the characteristics of discontinuity,high delay,high error rate and so on,and different from the TCP/IP network has a communication link from the source node to the destination node,which rarely exists stably in the opportunity network.A new method of "storing,carrying and forwarding" was developed in this network.After the relay node makes a connection with other nodes and successfully exchanges messages,it will first store the message in its own cache,then wait for the opportunity to meet other nodes to forward the message through the movement of the node,and so on,until it finally meets the destination node of the message.As a new network,the opportunity network is mainly used in some challenging special environments,such as wildlife tracking,communication in sparse environment in remote areas,on-board network and so on,which has a good development prospect and broad application space.Because the nodes in the opportunity network are constantly moving,which leads to the intermittent connection between nodes and the unpredictable encounter of nodes,how to select the appropriate relay nodes to make the message have a higher probability of transmission to the destination node is a key research direction in the design of routing algorithm.At the same time,in order to improve the probability of successful message transmission in the opportunity network,actively copy the message to the network,so that many redundant copies of are retained in the network,how to quickly delete redundant message copies while saving resources is another problem worth studying.After studying the existing routing algorithms,a probabilistic route based on utility value is proposed.At the same time,based on the research of existing message confirmation mechanism,a message confirmation mechanism based on node similarity in multi-copy environment is proposed.(1)A probabilistic routing algorithm based on utility values UV-Prophet,the historical encounter information of nodes has great reference value for whether they can meet in the future.Therefore,the encounter information of nodes is recorded in the process of node meeting with other nodes,and the forwarding utility value between nodes is derived by the duration of node meeting and the remaining cache of encounter nodes.The value of utility is used to calculate the transmission probability of the node,and the value of transmission probability and the threshold value are compared to determine the number of messages forwarded.If the transmission probability of the encounter node is greater than the node and less than the threshold value,a 1/2 copy of the message is distributed to the encounter node.If the transmission probability of the encounter node is greater than the threshold,the copy of the message will not be retained after forwarding.Comparison was made between UV-Prophet algorithm And Epidemic algorithm,Prophet algorithm,Binary Spray And Wait(BSW)algorithm in four aspects:cache size change,simulation time change,TTL,And message generation interval.Simulation results show that UV-Prophet can select the next hop more accurately,improve the delivery success rate,and reduce the transmission cost and average hop.(2)A message confirmation mechanism(SD-MCM)based on node similarity in multi-copy environment is proposed.The nodes in the opportunity network will show some stable characteristics after long-term contact,and the nodes that meet frequently will gradually form a stable social circle.Because of the similar activity range,the nodes in the same social circle are more likely to receive messages from other nodes in this range,so it is more likely to carry the same message copy.The node sends a confirmation message to other nodes in its social circle in a flood,so that the confirmation message can be spread in the social circle in time,so as to promote the circle to quickly delete the redundant copy of the message and release the cache space.The SD-MCM algorithm is compared with active and passive deletion methods in four aspects:cache size change,simulation time change,node number change and message generation interval.The simulation results show that the SD-MCM algorithm can avoid generating too many confirmation messages,causing extra cost,and is more efficient than the passive deletion method.