Comprehensive Weights And Punishment Mechanism Based On Trust-cloud In P2P Network

With the development and maturity of P2P technology, trust model has gradually shaped up and made an extensive use in P2P technology. As also, it has made great contributions for accurate and effective trust evaluation among the nodes in P2P network. In addition, it has an function of predicting the tendency of the whole network.The trust model of P2P network means to establish a mechanism which can quantify the trust and compute the trust degree of the node. In the model,every node sets a threshold of trust which used to measure the node it will interact with. If the trust degree of service-provided node is higher than the threshold,the service-received node will chose to believe it and interact with it. But if its trust degree is lower than the threshold, the service-received node will identify that it has a trust crisis and refuse to make a relationship with it. Similar to human society, subjective trust of P2P network has characteristic of uncertainty. However traditional method either did not take it into account or just considerer it as pure randomness or fuzziness which caused model imperfect.Professor Li who is an expert in the field of artificial intelligence proposed the concepts of the cloud model and discussed the randomness and fuzziness in-depth. Meanwhile, the cloud-model he proposed can process the accurate linguistic value into imprecise numerical value which made the result more precise.The cloud-model be applied to trust model give a full consideration of the changeability of the abundant nodes in the P2P network and the uncertainty caused by their diversity. Three digital characteristics have been introduced to combine the trust degree and the uncertainty that conform to the logic of humanity.The cloud-model also consider the possibility that two nodes would establish an trust relationship without any interactions ever since. The model introduce an method to aggregate several trust paths into one and make sure the aggregative path is authentic which reduce the risk of transaction. But this model did not consider the time factor and could not reflect the dynamic changing process. It also not have penalty strategy to punish the nodes who provide false service and make malice attack to destroy the environment. Therefore the cloud-model will turn into a more reasonable and robust model if considering the time factor and penalty strategy.A comprehensive mechanism base on cloud-model has been made in this paper which has fully consider the dynamic change of trust degree with the time. The transaction weights and the success rate have been taken into the model to response the trust status more precise. The numerous factors considered in human society have been brought into to make sure the model more comprehensive.Finally the punishing mechanism introduced in this model can punish the nodes that have provided dishonesty service or made malice attacks. With the mechanism, it turn to be a integrated model which can ensure the safety and stabilization of the network.Comprehensive model has a more abundant connotation in trust. At last we simulate a P2P network and introduce the trust-model into it. The result proves that the comprehensive cloud-model can bring a efficient and accurate trust evaluation in the P2P network.