基于行为模式的漏洞检测方法  

Vulnerability Detection Technology Based on Behavior Pattern

作　　者：葛学帅 刘铁铭[1] 谢耀滨[1] 张有为 GE Xueshuai;LIU Tieming;XIE Yaobin;ZHANG Youwei(Information Engineering University,Zhengzhou 450001,China;Zhengzhou Xinda Advanced Technology Research Institute,Zhengzhou 450001,China)

机构地区：[1]信息工程大学,河南郑州450001 [2]郑州信大先进技术研究院,河南郑州450001

出　　处：《信息工程大学学报》2024年第3期315-322,共8页Journal of Information Engineering University

基　　金：河南省科技攻关项目(222102210260)。

摘　　要：目前基于符号执行的漏洞检测工作仍存在漏洞类型有限和未能精确描述检测到的漏洞类型等问题。首先,通过深入分析常见内存损坏漏洞的形成原因和行为模式,提出了一种通用的基于行为模式的漏洞检测方法,覆盖更多的漏洞类型,并且在检测到漏洞的同时自动确定漏洞类型。最后,设计并实现了一个自动化漏洞检测框架,并通过实验验证了其有效性。研究成果对于提高网络空间安全能力具有一定的意义和价值。Existing vulnerability detection approaches based on symbolic execution still suffer from limitations such as a restricted range of detectable vulnerability types and imprecise characterization of the detected vulnerabilities.In this paper,we propose a generic vulnerability detection method based on behavioral patterns by analyzing in-depth the formation causes and behavioral patterns of common memory corruption vulnerabilities.This approach covers a wider range of vulnerability types and can automatically identify the type of the detected vulnerability.Furthermore,we design and implement an automated vulnerability detection framework and experimentally verify its effectiveness.Our study is of significance and value in improving network security capabilities.

关 键 词：漏洞检测 符号执行 内存损坏漏洞 

分 类 号：TN915.08[电子电信—通信与信息系统]