Keyless Public Watermarking for Intellectual Property Authentication

Abstract

A constraint-based watermarking technique has been introduced for the protection of intellectual properties such as hardware, software, algorithms, and solutions to hard problems. It provides desirable proof of authorship without rendering the intellectual properties useless. However, it makes the watermark detection, which is as important as watermarking, a NP-hard problem. We propose a keyless public watermarking method that enables the watermark to be publicly detectable. The basic idea is to create a cryptographically strong pseudo-random watermark, embed it into the original problem as a special (mutual exclusive) constraint, and make it public. No key is required to detect such watermark. We combine data integrity technique and the unique characteristics in the design of intellectual property such that adversaries can get almost no advantage for forgery from the public watermarking. This new technique is also compatible with the existing constraint-based watermarking techniques to enhance the strength of the watermark. We build the mathematical framework for the this approach based on the concept of mutual exclusive constraints. We use several well-known problems to explain our approach, demonstrate its robustness against attacks, and show that there is little degradation in performance.