摘要
密码函数识别在恶意代码分析、软件脆弱性分析等领域具有积极的作用。传统的密码函数识别算法由于识别方式单一而存在识别精度不高的问题。针对上述问题,提出一种改进的基于数据流分析的密码函数识别方法,将数据流分析引入密码函数识别中,利用递进式多特征的方法对密码函数进行识别。实验表明,该方法能够准确定位密码函数在应用程序中的位置,相比现有方法提高了密码函数的识别精度。
Cryptographic function recognition has a positive effect in malicious code analysis,software vulnerability analysis and other fields. Traditional cryptographic function recognition algorithm has the problem of low identification accuracy due to its single mode. In light of the above problem,we proposed an improved method of cryptographic function recognition which is based on data flow analysis. It introduces data flow analysis to cryptographic function recognition,and uses progressive multi-feature approach to recognise the cryptographic functions. Experiments showed that the method could accurately locate the position of cryptographic functions in the application,compared to existing methods,the accuracy of cryptographic function recognition was improved.
出处
《计算机应用与软件》
CSCD
2016年第3期299-303,313,共6页
Computer Applications and Software
基金
国家自然科学基金项目(61309007)
国家科技支撑计划项目(2012BAH47B01)
关键词
密码函数识别
常数特征匹配
数据流分析
Cryptographic function recognition
Constant feature matching
Data flow analysis
