Research On Dynamical Credibility Assessment Of Privacy-Preserving Strategy For Mobile Crowd Sensing

Mobile Crowd Sensing(MCS)is becoming a popular paradigm to collect information,which have the potential to change people's life.In MCS,participants upload data to the sensing platform and then complex social sensing missions can be executed though the intelligent data extraction.However,the system environment of the MCS is more open where nodes can voluntarily upload sensing data relying on auto-registered mobile networks and public networks.Due to the continuous data transmission process,the sensing data and node privacy are easily exposed to the relay nodes and channel monitors,which greatly increases the risks of hacker attacks and confidential leakage.Traditional privacy protection methods rely on encryption methods,not only consuming network resources,but also greatly affecting network performance.To tackle these issues concurrently,this thesis studied the privacy protection for mobile crowd sensing.In this thesis,the mobile crowd sensing network and the security challenges are introduced firstly,especially the privacy protection problems of task allocation and data transmission and detail the analysis of existing privacy protection strategies.Then,aiming to the privacy protection problems for coordinated task assignment,a trust status aware privacy protection strategy is proposed,which protecting the privacy of task requesters and participants concurrently.In this strategy,bloom filter is used as an attribute storage structure,and the participants' trust attributes and degree of preference are fully considered.Then,based on the privacy protection methods and historical information,the reliable participating nodes are selected.The effectiveness in privacy protection and efficiency of the proposed scheme are validated through numerical results.And then,aiming to the privacy protection problems for data forwarding,this thesis proposes a dynamical credibility assessment of privacy-preserving strategy.In this strategy,the sensing data are dynamically split into slices and the number of slices based on the trust of encountered nodes.Moreover,the secret sharing scheme and anonymous strategy ensure the data integrity and the anonymity of participants.Numerical analysis shows that under the premise of guaranteeing the network performance,the proposed strategy can control the leakage rate of data at about 6%.At last,we conclude our work and show our plan in future work.