QBDD: Quantum-resistant blockchain-assisted deep data deduplication protocol for vehicular crowdsensing system

Abstract

Vehicular Crowdsensing System (VCS) has emerged as a promising paradigm for alleviating traffic congestion and improving driving safety due to its convenient collection and aggregation of various driving and traffic-related reports. However, the rapidly growing number of vehicles means that VCS faces the severe challenge of large-capacity data storage. It has been observed that various data deduplication techniques have been devised for VCS. Unfortunately, in most solutions, the deduplication operations are centralized, unable to simultaneously reduce storage costs and bandwidth consumption. Moreover, these schemes are mainly constructed based on number-theoretic assumptions, which makes them susceptible to quantum attacks. Inspired by this fact, we introduce QBDD, a quantum-resistant blockchain-assisted second-level data deduplication protocol for privacy-preserving VCS, enabling efficient deduplication on the Road Side Unit (RSU) and Cloud Service Platform (CSP). We construct this protocol based on the Ring Learning with Errors (RLWE) problem signcryption and proxy re-encryption schemes to ensure data security. The improved Message Locking Encryption (MLE) mechanism hides the label of ciphertext and allows for quick duplicate crowdsensing report checks without compromising data privacy. Additionally, a blockchain-based distributed system is used to store crowdsensing data and maintain greedy lists, enhancing the system’s efficiency and fairness. We provide theoretical proof of QBDD’s security and conduct extensive experiments to demonstrate its practicality in post-quantum secure VCS.