Research on Data Privacy Protection of Internet of Vehicles Based on Differential Privacy

Abstract

Abstract The private data of intelligent vehicles connected to the Internet are easy to leak through the network. However, all existing solutions require the intervention of trusted third parties. In reality, even the confidentiality commitment provided by vehicle service providers is questioned by users. Aiming at this problem, a local differential privacy algorithm is introduced to design a data privacy protection scheme adapted to the characteristics of Internet of vehicles. The original data are only stored locally, and all the statistical data provided to the collector are disturbed. On the basis of ensuring the necessary data collection of the service center, it is guaranteed by mathematical principle that the collector cannot reverse the original data and the original data provider. The experimental results show that when the number of users is 10 million, the maximum relative error of frequency estimation is 0.0031 and the maximum relative error of mean estimation is 0.00006, which are far less than the tolerance value of 0.01, indicating that the scheme guarantees the application of Internet of vehicle safety situational awareness and other applications that need a lot of data support on the basis of protecting user privacy.