DESIGN AND ESTIMATION OF FEATURES BASED SOFTWARE BIRTHMARK

Software theft and piracy are rapidly increasing problems of modern day software \ud industry. This involves copying and reusing software without proper authorization \ud enforced by the software license agreements. Piracy of software ultimately results in big \ud financial loss to the copyright holder. General user community is not well aware of this \ud serious crime, and most of the time people think that it is not important for them to worry \ud about. Different advanced techniques are being used by practitioners to detect and \ud prevent software theft, including software watermarking and software fingerprints. But \ud the use of countermeasures, such as code obfuscation and optimization of software for the \ud semantic transformation of watermark, has made their use partially illogical. In the same \ud direction, a lot of research has been conducted to develop the concept of software \ud birthmark, which is now a widely accepted technique to detect software theft. A software \ud birthmark is based on the inherent properties of software. Accordingly, researchers have \ud proposed different categories and types of software birthmark based on some defined \ud attributes. Two attributes, credibility and resilience, are considered as the most important attributes of a software birthmark. Although the concept and types of software birthmark \ud is extensively studied by software developers and practitioners, there is still a lack of \ud knowledge and understanding about how to estimate a birthmark to analyse the extent of \ud piracy.The aim of this research is to promote conscious efforts made during software \ud development so as to incorporate well defined design features, resulting in software with \ud strong birthmark, hence preventing software piracy. For this purpose, a feature-based \ud software birthmark has been proposed which provides a closer and authentic \ud identification of software, and then ultimately be used for theft detection. This research \ud also presents a formal estimation model for features-based birthmark which can be used \ud to detect and investigate the extent of piracy in software. The model is tested through a \ud case study. Results exhibit correctness and usefulness of the proposed features based \ud software birthmark and its estimation model. A mathematical model is also presented, \ud through which the birthmark of software(s) can be compared for analyzing extent of \ud piracy