Privacy-preserving Scheme in Social Participatory Sensing Based on Secure Multi-party Cooperation

Ye Tian,Xiong Li,Arun Kumar Sangaiah,Edith Ngai,Zheng Song,Lanshan Zhang,Wendong Wang

DOI: https://doi.org/10.1016/j.comcom.2017.10.007

IF: 5.047

2017-01-01

Computer Communications

Abstract:Social participant sensing has been widely used to collect location related sensory data for various applications. In order to improve the Quality of Information (QoI) of the collected data with constrained budget, the application server needs to coordinate participants with different data collection capabilities and various incentive requirements. However, existing participant coordination methods either require participants to reveal their trajectories to the server which causes privacy leakage, or tradeoff the location accuracy of participants for privacy, thereby leading to lower QoI. In this paper, we propose a privacy-preserving scheme, which allows application server to provide quasi-optimal QoI for social sensing tasks without knowing participants' trajectories and identity. More specifically, we first suggest a Secure Multi-party Cooperation (SMC) based approach to evaluate participant's contribution in terms of QoI without disclosing each individual's trajectory. Second, a fuzzy decision based approach which aims to finely balance data utility gain, incentive budget and inferable privacy protection ability is adopted to coordinate participant in an incremental way. Third, sensory data and incentive are encrypted and then transferred along with participant-chain in perturbed way to protect user privacy throughout the data uploading and incentive distribution procedure. Simulation results show that our proposed method can efficiently select appropriate participants to achieve better QoI than other methods, and can protect each participant's privacy effectively.