Abstract: In Vehicular Ad-Hoc Networks (VANETs), existing location privacy protection schemes based on K-anonymity, mix-zones and path obfuscation all rely on the assistance among surrounding vehicles, which cannot effectively protect users' location privacy in a sparse vehicle scenario. To address this problem, a virtual location-based privacy protection scheme is proposed, in which a virtual location generation algorithm is designed and combined with K-anonymity techniques to achieve the obfuscation of the virtual location to the user's real location and reduce the probability of an attacker obtaining the user's real location. The experimental results show that the scheme can provide real-time and effective location privacy protection for users in different vehicle density scenarios, while reducing the tracking success rate to less than 1% in a shorter period of time.