In VANETs, vehicle users typically need to provide their actual location information to the RSU when requesting location-based services. However, there may be untrusted nodes around the Road Side Unit, and attackers may obtain the service data and location information of vehicle users through eavesdropping. Consequently, this results in the potential exposure of users’ location privacy. In response to this challenge, the present study introduces an intrusion detection-based location privacy protection algorithm. The algorithm employs Deep Belief Networks for data reduction to eliminate redundancy and utilizes the C4.5 decision tree for data classification to filter out malicious behavior data. The experimental results indicate that the proposed algorithm is capable of accurately distinguishing between attack data and normal service request data. While protecting user’s location privacy, it improves the quality of service, service request success rate, and service query accuracy.