SybilEye: Observer-Assisted Privacy-Preserving Sybil Attack Detection on Mobile Crowdsensing

Abstract

Mobile crowdsensing is a data collection system using widespread mobile devices with various sensors. The data processor cannot manage all mobile devices participating in mobile crowdsensing. A malicious user can conduct a Sybil attack (e.g., achieve a significant influence through extortion or the generation of fake IDs) to receive an incentive or destroy a system. A mobile crowdsensing system should, thus, be able to detect and block a Sybil attack. Existing Sybil attack detection mechanisms for wireless sensor networks cannot apply directly to mobile crowdsensing owing to the privacy issues of the participants and detection overhead. In this paper, we propose an effective privacy-preserving Sybil attack detection mechanism that distributes observer role to the users. To demonstrate the performance of our mechanism, we implement a Wi-Fi-connection-based Sybil attack detection model and show its feasibility by evaluating the detection performance.