Privacy-preserving and verifiable data aggregation for Internet of Vehicles

Abstract

With the development of Internet of Vehicles (IoV), organizations or institutions are increasingly interested in aggregating vehicle perception data to perform data analysis. Nevertheless, the existing solutions are difficult to meet the comprehensive requirements of IoV scenario, such as low latency, privacy preservation and rich functions. Additionally, it is more challenging that the aggregation node may return tampered or forged aggregation results for vulnerabilities or ulterior motives. To this end, we propose a Privacy-Preserving and Verifiable Data Aggregation Scheme for IoV (PPVDA), which allows the efficient aggregation for inner product over multi-dimensional data, and the data requester can verify the correctness and integrity of the aggregation results. By exploiting homomorphic MAC and secret sharing techniques, we construct a lightweight and verifiable mechanism for results as well as data privacy protection. We also conduct a comprehensive security analysis of PPVDA under the malicious security model. Moreover, extensive theoretical analyses and experimental evaluations demonstrate that PPVDA performs efficiently while retaining more desired properties. When dealing with 4×104 VNs, 3000 RSUs and 500-dimensional vectors, each EN side only takes about 74.5 ms and the DR side only takes about 3.8 ms, is therefore suitable for practical IoV scenarios.