Research On Identification Of Non-complete Cooperative Nodes In Mobile Peerto-peer Network

With the continuous improvement of the requirements for data transmission performance in the network,the research on Non-complete cooperative nodes has become more and more in-depth.The recognition of Non-complete cooperative nodes based on the trust model has also received wide attention,but the classification of Non-complete cooperative nodes research on identifying and non-complete cooperative nodes collaborating on identifying is very limited.To the mobile peer-to-peer network,on the one hand,due to its subjective rational selfishness,causes the node to generate a variety of uncooperative behaviors to avoid the consumption of its own resources.On the other hand,the limited objective resources make many functions of the node rely on the cooperation between the nodes to complete.This paper focuses on the classification and identification of non-complete cooperative nodes in mobile peer-to-peer networks,and tries to maximize the value of non-complete cooperative nodes so as to improve the efficiency of mobile peer-to-peer network data transmission..First,the concept of limit is introduced.The evaluation node(current node)analyzes the evaluated node from the three dimensions of the forwarding behavior,internal state and recommendation trust of the evaluated node(neighbor node),and realizes full coverage of the classified node being evaluated.Finally,the evaluated nodes are divided into four categories: cooperative node,lazy node,hypocrisy node,and malicious node,and they are defined.Calculate the forwarding rate of the evaluated node to quantify the forwarding behavior of the node being evaluated by listening in the hash listener mode the number of packets forwarded and the number of received packets;the concept of credibility is defined to represent the measurement of the authenticity of the recommendation information provided by the evaluating node to the evaluated node,and quantified by the evaluation node and the evaluated node respectively for their common neighbor nodes' forwarding rate and recommendation trust difference;to guess the internal state of the evaluated nodebased on the ratio of the number of unsolicited packets sent by the node being evaluated as the source node to the number of packets forwarded.By quantifying the three dimensions of the evaluated node,the categories of the evaluated nodes are identified,and then the trust model based on the identification of non-complete cooperative nodes is established,and the method of recommending trust and regularly updating the trust values reduces the effect of the node mobility on the trust model.Then on the basis of identifying single non-complete cooperative nodes,Collusion Non-complete Cooperative Node Identification Algorithm is proposed to help the evaluation node to identify the nodes in its communication range participating in the network collusion using the vector cosine similarity of the evaluated nodes for malicious nodes and hypocritical nodes,and using trend vectors for lazy nodes and cooperative nodes.Finally,according to the research on the identification of non-complete cooperative nodes,Non-complete Cooperative Node Identification Route Algorithm is designed.The relevant researches in this paper are all implemented on the NS2 simulation platform,using the six performance indicators of recall rate,false detection rate,accuracy,accuracy rate,error rate and rejection rate to evaluate Single Non-complete Cooperative Node Identification Algorithms and Collusion Non-complete Cooperative Node Identification Algorithm,compared with the Watchdog algorithm and the TQR algorithm.The performance of Non-complete Cooperative Node Identification Route Algorithm is evaluated using delivery rate and average end-to-end delay,and compared with AODV,Watchdog algorithm and TQR algorithm.In order to fully evaluate the effect of algorithm performance on the network,all performance metrics in this paper are measured using the average of all nodes in the network.The simulation results show that Single Non-complete Cooperative Node Identification Algorithms can accurately identify the type of the evaluated node;Collusion Non-complete Cooperative Node Identification Algorithm enables the node to better identify the members of the collusion group within its communication range in the mobile network environment;Non-complete Cooperative Node Identification Route Algorithm can make full use of the value of non-complete cooperative nodes,effectively avoid false recommendation information provided bynon-complete cooperative nodes,and can motivate nodes to cooperate and participate in data forwarding.It has good performance in data delivery rate.