Service similarity-based k-anonymity location privacy preserving method for Internet of Vehicles
Xuwei Ren, Zhexue Jin, Yongzhen Li · 2025
Aiming at the shortcomings of current location privacy protection methods for internet of vehicles (IoV) in terms of balancing location privacy and service availability, high communication overhead, and susceptibility to inference attacks, a k-anonymity location privacy protection method for IOV based on service similarity is proposed to effectively defend against inference attacks. The concept of service similarity is introduced to generate a service similarity map, and based on the service similarity degree, the partition where the vehicle is located is merged with other partitions to form the Anonymous candidate area that meets the service quality requirements of the vehicle. When selecting the k-anonymity set, the anonymity entropy is used to quantify the user's enquiry probability, and the k-anonymity set with the largest entropy is generated to effectively undercut the inference attack. A greedy algorithm is used to randomly select a location point in the anonymity set to request the service in order to reduce the resource overhead. And experimental results and analysis show that the method reduces the communication overhead by an average of 48.75% and improves the privacy preservation degree and service availability by an average of 42.99% and 45.84% compared to the comparison method. The proposed method improves the privacy preservation degree and service availability, reduces the resource overhead and is effective against inference attacks.