| dc.description | A fundamental concern in computer security is to control information ow, whether to protect con dential information from being leaked, or to protect trusted information from being tainted. A classic approach is to try to enforce non-interference. Unfortunately, achieving non-interference is often not possible, because often there is a correlation between secrets and observables, either by design or due to some physical feature of the computation (side channels). One promising approach to relaxing noninterference, is to develop a quantitative theory of information ow that allows us to reason about how much information is being leaked, thus paving the way to the possibility of tolerating small leaks.
In this work, we aim at developing a quantitative information ow C++ toolkit library, implementing several algorithms from the areas of QIF (more speci cally from four theories: Shannon Entropy, Min-Entropy, Guessing Entropy and G-Leakage) and Di erential Privacy. The library can be used by academics to facilitate research in these areas, as well as by students as a learning tool. A primary use of the library is to compute QIF measures as well as to generate plots, useful for understanding their behavior. Moreover, the library allows users to compute optimal di erentially private mechanisms, compare the utility of known mechanisms, compare the leakage of channels, compute gain functions that separate channels, and various other functionalities related to QIF. | |
