Researches And Applications Of Secure Set Operations Protocols

Secure Multi-party Computation(SMC)combines the cryptography and distributed computing technology.In a network computing environment where everyone does not trust each other,participants use their own private information to compute a function together.After the calculation,each party can get the correct output,but cannot get any information from the other parties.This problem can be solved by SMC,which makes itself a natural protection of user privacy.The secure computation of set is one of the hot issues in the research of SMC.Nowadays,most security computation schemes of set operations still have the problem of lower efficiency,security vulnerabilities or inability to securely store.In this paper,the set problem is transformed into a mathematical model problem by using set representation.Combining the cryptography tools,the secure computation protocols of set membership,set intersection and set intersection cardinality were proposed.The following three aspects are the specific content of this paper:First of all,we analyzed the security from the existing flaws of a recently efficient set membership computation protocol.After analysis,we found that the security vulnerability can obtain the element information input of the participant through brute force.To overcome this security vulnerability,based on the discrete logarithm problem,we proposed a new secure computation protocol of set membership by representing sets by polynomials.With the new protocol extending,we proposed a secure computation protocol of set intersection.At last,under the semi-honest model,we utilized the probabilistic polynomial time simulator to give the security proof of the new protocol.Moreover,we analyzed the performance of the new protocol in detail.The results are shown that the new protocol has better security performance and lower complexity.Secondly,we propose a new approach for sets representation,which denotes sets by bit vectors and naturally hides the cardinality of a set.The new approach is particularly suitable to the cloud computing environments.Then,we present the private set intersection(PSI)and PSI cardinality protocols based on an additive homomorphic public-key cryptosystem(PKC).The new protocols enjoy two main advantages:(a)they are more efficient than other related protocols especially when the set size is less than2¹²;(b)our approach provides a good solution to securely store users'datasets and the encrypted datasets could be used as protocols'messages directly without any additional computations.Finally,we implement our PSI and PSI cardinality protocols with Paillier PKC and ElGamal PKC in Java.Finally,we applied the pretty efficient PSI protocol we proposed to design and develop the network friend-seeking system of privacy preserving.Users can match and find friends with the same interests and hobbies on the premise of nondisclosure privacy information.Firstly,we analyzed the system architecture and the function module,and then elaborated the design and realization of system function.At last,we gave the detailed system function test and performance test analysis.