Preserving User Location Privacy Based on Web Queries and LBS Responses

Abstract

There is continued growth in online services that provide users with content based on location. These location based services (LBS) all require a location declaration by the user (or his electronic device). A typical example is a web query for services within an area in which the user is interested. Since the user cannot control the use of the data included in his query once it leaves the device or the response, he must assume that information is available to an unknown observer, thereby creating the potential for the user's location privacy to be compromised. This paper introduces the concept of a location privacy threshold which has two components: distance from the true location and probability that an observer would select the user's true location from a set of bogus locations. To preserve the user's location privacy, his location can be cloaked by adding additional web queries for the same information in areas near his designated location. A minimum spanning tree algorithm is used to identify clusters of potential user locations and determine whether the queries and replies provide enough uncertainty in the user's location to meet his predetermined privacy threshold.