基于密码隔离的恶意代码免疫模型  

A Crypto-based Immunization Model against Malicious Code

作　　者：陈泽茂[1] 沈昌祥[2] 吴晓平[1] 

机构地区：[1]海军工程大学信息安全系,武汉430033 [2]海军计算技术研究所,北京100841

出　　处：《计算机科学》2008年第1期288-289,293,共3页Computer Science

基　　金：国家“八六三”高技术研究发展计划项目(2002AA144020)的资助

摘　　要：基于访问控制的恶意代码防御模型对恶意代码实行的是逻辑隔离,它可能被旁路且防御效果受限于访问控制策略的有效性。本文基于密码学原理建立了一个恶意代码免疫模型,以克服逻辑隔离的脆弱性;定义了代码植入规则、保护规则和执行规则,实现代码存储和执行的安全;证明了在系统初态安全且代码加、解密密钥安全的条件下,任何时刻恶意代码都不会被执行和传播。Access control based malicious code defending model has vulnerabilities because it may be circumvented in certain circumstances. In addition, its effectiveness depends on that of the access control policy applied. A crypto-based security model against malicious code is proposed which defined secure operation rules for executable code planting, protection, and execution. It is proved that if the initial system state is secure and the crypto keys for code protection are secure also, the proposed model will be able to protect system from being infected and be able to prevent subjects from invoking malicious codes, thus to stop malicious code from spreading.

关 键 词：安全模型 恶意代码 安全操作系统 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]