A Software Watermark Scheme Based On The Instruction Encoding

The popularity of the Internet have promoted the rapid development of computer softwareindustry, however, it also brought a serious piracy problem, affecting the healthy development ofthe software industry, which raised the specter concern for software protection. Softwarewatermarking is a new software copy protection technology in recent years, it belongs to abranch of the digital watermarking technology, taking the software as the operation object, itembedded the copyright protection information which can be used to identify the author, issue,the owner, the use of those and other authentication information into software works. In theevent of copyright disputes, with the specific extraction algorithm, the embedded information isextracted to prove copyright attribution and tracking piracy. It is a powerful means of softwarecopyright protection, and has broad prospects for development, has developed a emergingdiscipline.Software watermark is divided into two kinds of static watermark and dynamic watermark.Currently it’s more mature in the study of dynamic software watermarking. Dynamic graphictechnology has the most potential for development. Most studies based on the source code orJava byte code, while better hidden and robustness of the machine code-based softwarewatermarking is the anther potential development of technical direction. In this paper, based onthe instruction encoding software watermarking is a native code implementation of technicalsolutions by analyzing to find the replacement instructions the encode it, read the sequence theyappear and change the sequence to achieve the watermark information extraction and embedding.When embedding, first to encrypt the watermark information and transform it to the binary form,disassembled software code segment, modify the software instruction that can be encoded in arandom position according to the specified encoding scheme to complete embedding thepretreatment watermark information, replace and modify the top40coded instructions of thesoftware to save the random location and pretreatment watermark’s length. The extraction is toread the disassembly of the top40coded instructions of the software to get the start position andlength, with them to achieve the watermark binary sequence which embedded in the software,after inverse processing and decryption can get the original software watermarking information.Analyzing the Windows executable files PE structure and the structure the of IA-32instructionformat, with the BeaEngine disassembly engine, we implement a program in the IA-32architecture and Windows operating system platform. In the practical level proved the feasibilityof the scheme. While these coded instructions used in the same instructions length and execution cycle, watermarked software’s size and performance properties without and change. The locationof embedded watermark is random and on behalf of watermark information instruction is theessential software code for the software, so resistant to many common attacks, and this codedinstruction are frequently in either RISC or CISC hardware architectures, so the program hascapacity embedded space, better hidden, strong robust and other advantages.