Research And Implementation Of P2P Network Reputation Data Storage And Recovery

With rapid development of P2P(Peer-to-Peer) network technology, P2P has already become an efficient platform for distributed applications depending on its advantages, such as dynamics, flexibility, robustness and self-organization, etc. It has widely applied to distribution computing, electronic commerce, file sharing and so on. As a large number of malicious nodes and anonymous reader access exist in P2P network, influence the operation of the node in a P2P network is seriously affected by malicious behavior, so trust mechanisms has become a major guarantee and efficient mechanism of the node collaboration. In a P2P network, data storage method mostly rely on the authoritative third party and super node, the restore for data can be done by restoring backup mechanisms, the security of data storage is an important safeguard for reputation mechanism to execute steadily in a P2P network. If the failure of multiple connected nodes will lead to system crash or the degradation of system performance, and it is very important to restore confidence in the reputation system. Therefore, how to ensure the safety and integrity of reputation data currently distributed storage and node faults must consider the question.In this dissertation, a series of key-techniques are researched such as trust model for P2P network and the storage for the credits data and the recovery for the credits data etc. The trust model can help to solve a lot of problems, like free riding and tragedy of common in a P2P network, but the existing model of P2P nodes do not take into account that recovery of the credits data makes the problem of how to storage and recovery data is very important to P2P network. However, the data storage and recovery are closely related to storage space and confidence of nodes. The biggest drawback of P2P network is that nodes have different storage spaces and confidence. The smaller the space, the precious storage ability is. It is very important that how to weigh the relations between storage space and confidence of nodes has been an important issue.Therefore, the above problems, firstly, a model integrating the node weight and the storage cost are proposed in this paper. Secondly, it focus on the problem about credibility of data storage and recovery. Finally, it elaborates on the detail design and implementation of the main function modules. For data storage and recovery, this paper combines DHT storage solution and a dynamic secret sharing. The Scheme firstly builds a distributed hash table based on the physical topology of network, then, according to the secret sharing scheme to calculate credits data share and store the result back in the node of Hash ring distributively, even if losing credits data share, the credit information is safe, the study shows that the proposed scheme can well prevent the credits data from tampered by malicious nodes. However, when recovering the credits data, firstly, reliable nodes are retrieved by query. Secondly, using secret sharing scheme to solve the issue of recovering credits data. At last, the simulation of the method was carried out and compared the result with the original method. It can verifies the effectiveness and the feasibility of the method.