Research On Privacy Preserving Of User Attributes Matchmaking Based On Multi-relational Mobile Social Network

Mobile social networks(MSN) have become an inseparable part of our lives with the wide application of personal hand-held mobile devices. MSN provides a platform of communicating with friends, posting states and expanding communication circle. Meanwhile, different MSNs start to interact together, one user may not register only one account in MSNs, different accounts also have some common properties. It can help providing personalized service to study on the common properties. On the other hand, there exists lots of security challenges in multi-relational MSN, it threatens users’ information security. Recent secure matchmaking protocol is mainly based on complex math theorem, so it is not so much efficient, and there is not framework for multi-relational MSN. It has very big significance to do research on protocols and framework adapt to multi-relational MSN.This thesis analysis the current situation and security problem of MSN, and research on the matchmaking protocols, then propose an efficient secure matchmaking protocol and a private attributes matching framework for multi-relational MSN.This thesis’ s main research contents include:First, it proposes an efficient and secure matchmaking protocols based on the distance between dots in geometric space. This protocol can find the intersection without disclosing private attributes except the common part. This protocol combines matchmaking and geometric computing based on dots’ distance computing and one-way hash function. At last, we analyze the correctness and security.Second, we design an efficient and secure matchmaking framework for multi-relational MSN based on our protocol and the features of multi-relational MSN. After introducing the framework, we list every steps of matchmaking under the framework. At last, we analyze the security of border server and users.Third, at the end of this thesis, we simulate the secure matchmaking protocols comparing with other three protocols and verify the correctness of protocol. Otherwise, we use Peersim to simulate the matchmaking process from response time and accuracy.