Intelligent Anti-Tamper Protection for Software

"Automatic code obfuscation is the most viable method to protect software against reverse-engineering attacks. However, current obfuscators apply obfuscation transformations haphazardly. The actual utility and impact of all accumulated transformations arenot considered. Instead, protection is hoped to arise as the number of complex transformations builds up.Yet, dependable software protection requires a trustworthy approach with guaranteed protection levels, and without incurring excessive and unnecessary computational overhead. Formal models for software protection are believed to offer such benefits (akinto guarantees in current cryptographic models), but no such models have been identified.An alternative route is Artificial Intelligence (AI). High-level planning and reasoned decisions can selectively guide application of transformations, while gauging complex trade-offs between guaranteed protection level and computational overhead. AImeta-cognitive planners are known for excelling in this type of situation, but were not applied to this particular problem.We propose to undertake this promising route, with the goal of producing an intelligent anti-tamper protection system for software that is effective, reliable, and trustworthy. The proposed system will benefit the software industry and the military. Preventing IP theft increases the ROI on software, and protects the United States technological military advantage.The system is to b