Structured Watermarks for Structured Software

Software watermarking involves integrating an identifier within the software, enabling timely retrieval to disclose authorship/ownership, and deter piracy. Various software watermarking schemes have been proposed in the literature, many of which involve statically embedding an encoded identifier into the control flow graph of the program. In this paper, we propose novel embedding and extraction algorithms characterized by four key features: randomization, generating watermarks with a size closely matching the number of bits in the identifier, implementing both encoding and decoding with linear time complexity, and, most importantly, generating watermarks that conform to structured code. We emphasize the capability to encode the same identifier as distinct graphs, coupled with the absence of cumbersome “GOTO”-like substructures, as enhancements to the stealthiness of our watermarks. This feature makes them more resilient to common forms of attack, contributing to their effectiveness in safeguarding software integrity and discouraging unauthorized use.