Privacy-Preserving Algorithm for APPs in Vehicle Intelligent Terminal System: A Compressive Method

Chenlu Gao, Jianquan Lu, Jungang Lou, Yang Liu, Wenwu Yu

Abstract: Vehicle intelligent terminals often require end-user applications (APPs) to continuously send information to external data aggregator for monitoring or control tasks. However, communication channels exposed to the open environment can lead to an undesirable loss of privacy for drivers, despite the benefits provided by these APPs. It is worth mentioning that different APPs of vehicle intelligent terminals may require different key information. Given these personalized application scenarios, the use of traditional privacy protection technology may encounter limitations. To address this problem, we propose a compressive privacy-preserving algorithm that employs a compression matrix to maintain the normal function of applications while ensuring privacy. Specifically, the algorithm utilizes sensors to linearly transform the original vectors of measurements at each time step into a lower-dimensional space. The compressed measurements are then transmitted to a fusion center. Optimization problems are formulated for the current time step as well as two time steps. The performance of the algorithm is evaluated using the Cramer-Rao bound. Simulation results and comparisons further validate the effectiveness of the compressive privacy algorithm.