Verification of Quantitative Information Flow

我们解决了开放的问题，各种定量信息流验证问题的复杂性。基于信度和最小熵信道容量等信息论概念，给出了信息流的定量定义，并从计算复杂性理论和程序验证属性分类两个方面对该问题进行了研究，给出了程序验证属性分类的“超属性”概念。“我们还提出了利用软件模型检查和计数算法精确推断和验证定量信息流边界的算法。我们还提出了新的软件模型检测方法。

We solved open problems concerning the complexity of various quantitative information flow verification problems. We considered quantitative information flow definitions based on various information theoretic notions such as belief and min entropy channel capacity, and studied the problems both from the computational complexity theoretic aspect and the program verification property classification aspect formalized by the notion of "hyperproperties." We also proposed algorithms for precisely inferring and verifying the quantitative information flow bounds that utilize software model checking and counting algorithms. We also proposed new approaches to software model checking.