|
P2P network has become one of the Internet an indispensable application , but , because of its dynamic , distributed , autonomous , and other features that make the node malicious behavior , such as unreliable service , a lot of fraud , no contribution to the network , resulting in node lack of trust between restricted the further development of P2P . So how to efficiently manage the trust between each node in the network has become an important subject of P2P network research . Establish a reliable trust model is an effective method to solve the problem of P2P network trust . No initialization , existing models can not prevent malicious behavior of nodes , the recommended route to find the low efficiency problem , this paper presents an improved trust model with rewards and punishments and incentives to deal with the trust between nodes in the P2P network the relationship between the design of the overall structure of the model , in-depth study of trust initialization process , the calculation of the trust , recommended trust path queries, and updates four aspects of the trust relationship . Trust ( 1 ) model initialization process , and to establish a direct relationship of trust table in each node local , distributed management of the trust ; (2 ) the calculation of the trust is considering direct trust , recommendation trust and credibility can be more correctly assess the relationship of trust between the nodes , and the introduction of the concept of time decay factor that trust between nodes with a time of growth and decay , the recent interaction much more effect on the trust relationship between the nodes ; (3) Recommended path query , design a search algorithm based on ant colony algorithm recommended path to prevent a malicious node joint fraud , and the introduction of incentives in the search process , encourage each node Recommended ; (4 ) in the trust relationship the introduced the incentive algorithm update node trust relationship between an update , a better solution to the security problems brought about by the node malicious behavior . Finally, the model of the safety analysis , the effects of computing and incentive trust update algorithm and trust , analysis shows that the trust model can effectively prevent the malicious behavior of nodes in the P2P network , and to improve the the recommended route searching efficiency .
|