Matlock: A location obfuscation technique for accuracy-restricted applications

Location-based information systems (LBIS) have shown growth in the last years thanks to the popularization of cheap GPS-enabled mobile devices with connectivity to the Internet. At the same time, users have made their locations more available than ever, representing a major privacy problem, as attackers could easily determine their victims' common paths or actual locations. As a result, some techniques have been designed to protect the users' private information while allowing Location Based Service systems (LBS) to continue growing. Location obfuscation is a technique that alters, substitutes, or generalizes the real location of the users to protect them against malicious attacks. Moreover, there are activities that require a high level of accuracy while also maintaining the protection of the location, such as fleet management or transportation of valuables. In this work, a light-weight and fully reversible location obfuscation technique named Matlock is presented. Matlock, based on matrix obfuscation, requires low computation per operation and hides the location in both spatial and temporal dimensions. The performance evaluation shows how this technique alters the location of paths from different geographical regions, produces a non-morphologically similar path to the original path, and generates increasing distance functions between the original and the obfuscated paths, which together should reduce the probability of using estimation techniques to perform deobfuscation of the information.