A New Method of Software Clone Detection Based on Binary Instruction Structure Analysis

This paper aims to propose a new method based on binary instruction structure analysis to detect the degree of similarity among softwares, which provides a technical reference for the software clone detection. Without source codes of software, it makes a reverse analysis of the disassembly of an executable file, and has similarity intercomparison through function snipping, instructions classifying and weighting on the function similarity which is computed by the longest common subsequence (LCS). In the process of intercomparsion, this paper has considered that the program execution process and the function call sequences have impact on the code layout. The algorithm is optimized to achieve more accurate similarity by parsing and optimizing the executable file structure on the level of instruction executive structure. It provides a new way to detect software clone in case of no source codes, which is proved effective by the experiment.