A Location Usability Measure Method for Distribute K-Anonymity Privacy Protection in LBS

Distributed K-anonymity is a well-known method for protecting location privacy in location-based services. In this method, the request user constructs an anonymous cloaking region with the real locations of cooperative users and then submits this region to location-based service provider (LSP). However, when directly adopting the existing schemes, the malicious cooperative users do not share their real locations with the request user but instead provide fake locations. If the request user employs these fake locations to construct the anonymous cloaking region, LSP can effectively reduce the anonymous cloaking region or even directly identify the real location of the request user. Therefore, the existing distributed
$K$
-anonymity schemes cannot completely protect the request user's location privacy. To solve this problem, based on the historical location dataset, this paper filters out the fake locations that can be identified by taking into account three aspects, namely visited time, visited distance, and visited frequency. In this way, the constructed anonymous cloaking region can be used to protect the request user's location privacy successfully. Extensive experiments indicate that the proposal is effective and efficient.