Vehicle Road Privacy Recommendation System Equipped with Markov Chain

Abstract

Due to the development of in-vehicle network, location-based service has brought many conveniences to users. However, the user’s behavior of constantly updating the location of the service provider will cause the private information to be exposed to the attacker, thus threatening the user’s information security. Most of the current schemes ignore the differential protection of different road environments, which may lead to the abuse of privacy budget or be excluded by attackers according to actual environment analysis. We propose a method to protect the road privacy and meet the individual needs of users, and carry out the differentiated privacy protection based on road environment. First, we calculate the length of each road section, create a normalized matrix by a Markov model to describe the congestion degree of road sections, and filter the best route according to users’ privacy preferences. Then, according to the congestion degree of the road section, the sensitive circle range is defined for each query position of the recommended route, and the acceptable deviation range of the user virtual position is superimposed to protect the differential privacy. Experimental results show that, compared with the current methods, the proposed scheme can reasonably protect users’ privacy and obtain better service quality while satisfying users’ preferences and considering the surrounding road environment.