Privacy-Preserving Location-based Services for Autonomous Vehicle Navigation

Abstract

In the field of autonomous driving, innovative location-based services (LBSs) with respect to real-time dynamic verification and correction assistance are becoming increasingly important, which could assist a vehicle in navigating through an urban environment safely and efficiently [1]. However, vulnerabilities in fundamental communication processes can lead to a wide range of dangerous attacks on the vehicle and threaten personal data privacy, and at present, no work has addressed location privacy for these data query mechanisms, especially in a vehicle-to-everything (V2X) environment. We consider this to be a significant challenge. In a vehicular ad-hoc network (VANET) architecture, vehicles communicate with other vehicles or roadside units (RSUs) over wireless channels and, to some extent, expose themselves to a set of security vulnerabilities. Therefore, they are not suitable for transmitting the user’s real-time location to the destination server.

Three types of secure location-based data query schemes exist: pseudonyms, location cloaking, and cryptographic techniques [2]. Pseudonym-based solutions disrupt the connection between vehicle identities and locations but may still disclose data query locations. Location cloaking introduces uncertainty or error into location information, but may not guarantee enough users in the neighborhood or high query accuracy. Cryptographic techniques effectively preserve query accuracy and location privacy. In this paper, we propose a privacy-preserving range query (PPRQ) scheme for the autonomous vehicle navigation scenario and provide security analysis and performance evaluation. The proposed PPRQ scheme supports data query for locations in both the User and Data models and achieves the tradeoff between location privacy and data query efficiency, by providing slider window privacy-preserving functions on the road network map. It preserves the location privacy while enabling efficient data query. The proposed real-time privacy-preserving range query scheme is suitable for autonomous vehicle navigation and has great potential to be extended to other privacy-preserving data query schemes and to promote the development of privacy-preserving location-based services [3].