Research On Routing Algorithm Based On The Combination Of Historical Meeting Probability And Gray Prediction Model

Current opportunistic network routing and forwarding policy transmission performance is relatively low.Common routing forwarding strategies are:Epidemic,Spray and Wait,and Prophet.The above three routing forwarding policies have the following problems:low message delivery rate,long average delay,and high node congestion rate.In view of this,the F-HGRA routing algorithm proposed in this paper solves the above problems.The research work and achievements of this paper are as follows:In Spray and Wait,the node that carries the message delivers the message to the target node with some degree of blindness.Prophet only considers the probability of encounters between nodes in the shortest period of time,and ignores the variation of the probability of encounters between nodes in each time period.In order to solve the deficiencies of the above two routing algorithms,an F-HGRA routing algorithm is proposed.In F-HGRA,the Flooding phase is based on routing policies that limit the number of message redundant copies.At this stage,L messages are randomly distributed to L randomly-moving nodes and the meeting information between the nodes is collected.Based on the meeting information,the probability of encounter between the nodes is further obtained.In the HGRA stage,the metabolic gray prediction model is combined with the residual correction model to model the probability of encounter between nodes collected during the Flooding phase to obtain a set of high-precision analog value sequences and the predicted probability of the next time unit.It shall to get the probability of encountering the forwarding node i,the encounter node j and the target node d respectively,and get the expected values E_id>E_jd,.The size of the two will be compared as a preliminary screening condition.The prediction value p_id^?m+1?and p_jd^?m+1?of the meeting probability of i,the meeting node j and the target node d in the next time are further obtained by the Grey Prediction Model,and the size of the prediction p_id^?m+1?*C andp_jd^?m+1?are compared with that of the target node.C is a parameter that is related to the proportion of cache space in node j.If E_id>E_jd,,it shows that the average meeting probability of the forwarding node i and the target node D is higher than the meeting node j,then give up the forwarding;otherwise,further compare the size betweenp_id^?m+1?*Candp_jd^?m+1?,if p_jd^?m+1??29?p_id^?m+1?*C,then forward to the node j;otherwise,give up the forwarding.Because c reflects the congestion level of the meeting node j.If the C is large,it puts forward higher requirements for the prediction value p_jd^?m+1?,so the congestion degree of the meeting node j is well controlled.Based on the detailed description of the F-HGRA routing algorithm and the Opnet simulation software,the F-HGRA is compared with the classical route algorithm of the opportunistic network.It is proved that the routing strategy not only improves the message delivery rate,but also reduces the average message transmission time.The delay reduces the node's congestion rate and the possibility of network congestion.