Specifying API trace birthmark by abstract interpretation

摘要

API trace birthmark is a major class of software birthmarks, where API sequences are defined as software birthmarks to detect software theft. Currently, many birthmarks of this class have been proposed, but the evaluation of these birthmarks are mainly done through experiments and there is no theoretical framework, which makes it difficult to formally analyze and certify the effectiveness of the birthmarks. To solve this problem, an abstract interpretation-based method for specifying API trace birthmark is proposed in this paper. First, API trace birthmark is characterized as a semantic program property by abstract interpretation. Then, the credibility of API trace birthmark with respect to a specific criterion for copy relation is formally analyzed. Finally, the resilience of API trace birthmark is discussed and it is proved that API trace birthmark is resilient to a commonly used program transformation.