Research On Applications And Protocols Of Secure Multi-party Computation

In the open network, information exchange between users is actually distributed computing, each user according to his requirement to complete their computing tasks, and they work together to get the final results. During the communication process, the user is directly involved in the operation, it may lead to private information stolen by others. It not only reveals our own confidential information, but also may cause serious consequences.Howtoensure the security of privacy data during network communicationis concerned to everyone. In order to solve the computing network security problems, secure multi-party computation(SMPC) came into being.Nowadays, the research of secure multi-party computation is divided into two areas: firstly, the theory part that is based on secure multiparty computation. It includes the concept, known results, security models, key technologies, programs and other construction-related research; the second part is the practical applications that are based on the secure multi-party computation protocols, which includes the protocols such as polls online, electronic auction,outsourcing services and so on. Although the concept of secure multi-party computation has some progress and achievements, however, some results do not meet the needs of practical application problems. Therefore, constructing the practical scenario, that achieves the needs of reality safe, high efficiency, and practical protocols, is a crucial issue for a great number of experts and scholars. This thesis gives the models and applications of secure multi-protocol, which is as follows:(1)We introduce the description and notion of secure multi-party computation and its research background, significance. We also analyze the state-of-the-art of research.(2)We propose a privacy-preserving information matching protocol in secure two-party computation that is based on mobile terminal to seek the friends in our current social network environment. We give the execution of the protocol in semi-honest model, as well as its security, accuracy, and complexity. The proposed protocol is effective and practical in when we making and seekig friends with others in open networks.(3)We alsopropose a privacy-friendly reputation weights aggregation protocol in cloud computing services. We give the execution of the concrete protocols both in semi-honest model and malicious adversary model. Meanwhile, we analyzes their security, accuracy, complexity etc.The protocol only needs two rounds of communication that is more efficient.Also our protocol can resist against at most n-2 malicious attackers. Compared with the other protocols, the proposed protocol has better performace.