Research On Privacy-Preserving Reputation Management Scheme For Blockchain-based Mobile Crowdsensing

Recently,with the popularization of mobile smart devices and the rapid development of Io T technology,mobile crowdsensing(MCS)has been an emerging data collection paradigm that exploits the potential of individual mobile devices to acquire mass data in a cost-effective manner.MCS now is widely used in traffic monitoring,environmental sensing and smart city.One of the most significant challenges in these applications is to resist malicious users who providing false data to disturb the system.In the existing work,the reputation management scheme is an effective way to overcome the challenge of malicious user.However,most reputation management schemes rely on a semi-honest server,which cannot guarantee security and reliability of the server.Moreover,the existing schemes usually process data in the plaintext domain without protecting users' sensitive information privacy.Therefore,it's clearly necessary to build a secure and privacy-preserving reputation management scheme to deal with the challenge of malicious user and privacy protection.To solve these problems,this paper does the following work in the MCS scenario:(1)This paper integrates the blockchain technology and edge computing to build a reliable and efficient MCS framework,called BC-MCS,overcoming the problem of the opaque and unreliable server in traditional MCS applications,implementing an open and supervised MCS system.(2)To resist malicious users,this paper utilizes secret sharing based on the features of BCMCS framework to present a privacy-preserving reputation management scheme.Specifically,in the sensing stage,secret sharing is used to split secret key realizing the secure data aggregation;in the reputation management stage,setting up impact factors off-chain to evaluate data quality and using secret sharing to assembly evaluation on-chain.This scheme achieves the goal of resisting malicious users and protecting the honest user's private data.(3)To solve the dynamic user problem in the BC-MCS reputation management scheme,this paper designs the delegation protocol to improve the basic scheme,which enhances the scheme's robustness and eliminates the off-line users' impact on the effectiveness and security of the scheme.(4)This paper implements the system prototype based on Hyperledger Sawtooth,evaluating the performance and effectiveness of the BC-MCS reputation management scheme in real data set and different situations.The result shows that the scheme in this paper can achieve higher utility and security levels in handling malicious users,compared with the existing reputation management schemes.