Sensitivity analysis and countermeasures for transformation-based location obfuscation

Development of location privacy protection mechanisms is one of the most important concerns in existing and future location-based information systems, due to the important value that users' locations have for ill-intentioned organizations that could break in and obtain this data. Transformation-based location obfuscation techniques are one of the best techniques available because they substantially alter the location of the user, without revealing the whereabouts of the routes, while being completely reversible. One example of this is the Matlock technique. Even though it has been proven that theoretically it can be reversed by solving a linear system, it is not clear to what extent one can feasibly obtain the original route or enough information to recreate it. This work performs an analysis on how sensitive Matlock really is to reverse engineering and how some countermeasures can decrease the accuracy of the estimations from a third party. The use of noise-based obfuscation is shown to be a useful tool to reduce the accuracy of the obfuscation matrix estimation.