Non-Inference: An Information Flow Control Model for Location-based Services

This paper presents a framework for preserving location privacy without affecting location accuracy. In this framework, services migrate a piece of code to a trusted server, which is assumed to have location information of all the interesting subjects. The code executes on the trusted server, reads location information and sends back results. We introduce non-inference, a novel information-flow control model that guarantees that the code does not leak exact location information. We discuss the design, implementation and evaluation of a static program analysis technique that enforces non-inference for location based services.