A Method for Generating Confusing Positions to Resist Long-term Observation Attacks

Abstract

With the upgrading of mobile smart devices and rapid innovation of positioning technology, the widespread application of Location Based Services (LBS) has been propelled. However, users face the threat of privacy leaks while using LBS, especially when their behaviors over a long period of time are collected and stored. The accumulated information can be exploited by adversaries, greatly increasing the risk of privacy leakage. This type of attack is known as long-term observation attack. How to better protect users' location privacy remains a challenging issue. In this paper, we propose a method for generating obfuscated locations to counter long-term observation attacks. Firstly, we determine the output set based on the user's quality of service requirements and custom sampling probability. Then, we introduce an improved quadtree structure to store user location information and select users with similar destinations to form an anonymous set. Finally, noise is added to prevent long-term observers from accurately obtaining the user's precise location. Experimental results demonstrate that our method can effectively protect users from long-term observation attacks while maintaining low time cost and high anonymity efficiency.