Research And Implementation On Software Watermarking Based On Permutation And PPCT Mixed Encoding

Since digital products can be duplicated and distributed easily and economically, software security and piracy becomes a more and more important issue. According to the statistics in 2001, software and technology piracy is suspected to cause approximately $16 billion business and increase rapidly. Software has penetrated almost all of the aspects of industry, agriculture, commerce and daily life nowadays, so various technologies for software copyright protection have been studied. So far, these technologies are based on cryptography or steganography. There are three kinds of technologies about this issue in steganography, Code Obfuscation, Software Watermarking and Tamper-Proofing. This thesis is concerned about Software Watermarking.Software watermarks can be classified by the way of embedding and extracting techniques as static or dynamic. Although it can be embedded and extracted easily, static software watermark can't be applied to the situation which needs robust watermarks because of vulnerability. On the contrary, dynamic software watermark is robustness, so it becomes a hotspot in this field. Dynamic Graph Watermark (DGW) is one of the best dynamic software watermarks. The idea of DGW is encoding the watermark into topology graph in order to be immune to semantics preserving attack. After user input the extraction key, the topology graph is reconstructed on the heap, and then the decoding algorithm retrieves the watermark. So DGW is more robust and stealthier than other software watermarks.The core issue of DGW is encoding and decoding with topology graphs. Currently, the robustness of Planted Plane Cubic Tree (PPCT) is strongest among the four basic graphs for DGW, but its data rate is lowest. For improving data rate of PPCT, a novel scheme which combines Permutation Encoding Enumeration and PPCT Encoding Enumeration is proposed in this thesis. The new scheme makes use of the structure advantage of PPCT and the high data rate of Permutation sufficiently.There are two methods for implementing the new scheme in Java, Exchanging LeafNodes Mixed Encoding Enumeration and Adding Pointers Mixed Encoding Enumeration. The former exchanges the positions of leaf nodes of PPCT for Permutation Encoding Enumeration, so its structure is similar to PPCT and robustness is strong. The latter adds pointers among the leaf nodes of PPCT to achieve the purpose, and the process of encoding is facile, moreover, the encoding time is short.