缓冲区溢出攻击的动态检测与防范  

Dynamic detection and prevention of buffer overflow attacks

作　　者：张楠[1] 张建华[1] 傅春常[1] 李建阳[1] 

机构地区：[1]西南民族大学计算机科学与技术学院,成都610041

出　　处：《西南民族大学学报（自然科学版）》2007年第5期1172-1174,共3页Journal of Southwest Minzu University(Natural Science Edition)

摘　　要：缓冲区溢出是被黑客利用得最多的漏洞之一,为了对其进行动态的检测及防范,提出了一种基于地址认证的缓冲区防护(Authentication-Guard)技术.该技术通过修改编译代码,动态检测缓冲区是否发生溢出,对返回地址进行保护,避免了恶意代码的运行.利用该技术,攻击者无法利用计算缓冲区长度来绕开检测,同时也避免了返回地址不可预知的问题.Buffer overflow is one of the holes that are usually used by hackers. In order to dynamically detect and prevent it, anauthentication guard technique is brought forward. The compile code is modified to judge whether the buffer overflows is happens, and prevent the hacker＇s program running. While this technique is used, the attacker can not steer clear of detection by counting the length of the buffer. At the same tine, this technique solves the problem that the return address is unknown.

关 键 词：缓冲区溢出 编译器 地址认证码 返回地址 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]