基于非线性模糊矩阵的代码混淆有效性评估模型

Code Obfuscation Effectiveness Assessment Model Based on Nonlinear Fuzzy Matrices

为了解决目前代码混淆评估方法对代码混淆效果区分度不高的问题,文中提出一种基于非线性模糊矩阵的代码混淆有效性评估模型MNLFM(Code Obfuscation Effective Assessment Model Based on Nonlinear Fuzzy Matrices),并证明了MNLFM具有评估合理性、单调递增性、连续性和突出性等特性。MNLFM可以明显改善当前代码混淆评估领域在混淆效果方面可区分性差的现状。通过量化评估指标、确定隶属函数和构造非线性模糊矩阵等方法进行建模。建立一个Java程序测试用例集,基于压扁控制流和多种不透明谓词代码混淆技术对此模型进行混淆有效性检验,并将其与其他代码混淆评估模型进行比较。实验结果验证了MNLFM可以比较混淆后代码之间的综合复杂度,并明确区分不同混淆算法对原代码的混淆程度。

In order to solve the problem of present code obfuscation assessment method for low level of the code obfuscation discrimination,this paper proposed a code obfuscation effectiveness assessment model based on nonlinear fuzzy matrices(MNLFM),and gave a proof of several MNLFM’s features,such as assessing rationality,monotonicity,continuity,highlighting.MNLFM can obviously improve the current situation of poor distinction in the field of obfuscation assessment.The model can be carried out by quantifying the assessment index parameters,determining the membership functions and constructing the nonlinear fuzzy matrices.A test case suite of Java program was set up and several code obfuscation technologies based on flatten control flow and opaque predicate were used to check the validation of the model.And then it was compared with other code obfuscation assessment models.The experimental results verify that MNLFM can compare the comprehensive complexity between the obfuscation codes and clearly distinguish the degree of different obfuscation algorithms for original code.