A Fog-Aided Privacy-Preserving Truth Discovery Framework over Crowdsensed Data Streams

With the proliferation of mobile and wearable devices, mobile crowdsensing (MCS) is becoming a new paradigm for data collection and analysis. To effectively identify truthful information from crowdsensed data without privacy leakage, privacy-preserving truth discovery (PPTD) has gained much attention recently. Existing works either didn't consider real-time applications over data streams or failed to achieve enough efficiency for a large group of workers. In this paper, we propose FPTD, a Fog-aided Privacy-preserving Truth Discovery framework which is secure and efficient in handling real-time applications with a large group of workers. To reduce overhead, we adopt cloud-fog computing architecture to divide the complete worker group into many smaller ones. Then we design a unique secure aggregation protocol SecAgg which can securely and efficiently aggregate inputs from workers in smaller groups. Finally, we give detailed construction of FPTD, an efficient truth discovery framework based on SecAgg for real-time applications. Through extensive experiments and security analysis, we demonstrate that both SecAgg and FPTD are secure and efficient.