树突状细胞分化模型在人工免疫系统中的应用研究  

Research on Differentiation Model and Application of Dendritic Cells in Artificial Immune System

作　　者：倪建成[1,2] 李志蜀[2] 孙继荣[2] 周利平[2] 

机构地区：[1]曲阜师范大学计算机学院,山东日照276826 [2]四川大学计算机学院,四川成都610065

出　　处：《电子学报》2008年第11期2210-2215,共6页Acta Electronica Sinica

基　　金：国家自然科学基金(No.60072014);四川省科技公关项目(No.05GG021-003-2);山东省自然基金(No.Q99G03)

摘　　要：树突状细胞(Dendritic Cell,DC)是先天性免疫系统的重要组件,其分化机制是正确引发与调节适应性免疫响应的关键.首先,在描述DC分化的生物机理基础上,抽象出了DC的信息处理过程.其次,在阐释DAMP等四类外部信号的含义与功能、信号融合过程的基础上,定义了未成熟、完全成熟与半成熟DC Agent,刻画了它们的分化数学模型与演化过程.最后,论证了各类DC Agent数量与生存周期之间的关系.实验结果表明DC分化机制对降低入侵检测误报率、实现自我调节和进一步增强计算机系统安全具有重要的理论意义与应用价值.The most one critical component in innate immnne system is dentritic cell （DC）, which differentiation mechanism is also the key to initiate and control adaptive immune response correctly. Firstly, based on describing biological principles of DCs differentiation, the information processing procedure for DCs is abstracted. Secondly, laying the foundation of illustrating four-cate- gory behavioral signals such as danger-associate molecular patterns et al., and informational fusion procedure, agents including immature, fully mature and semi-mature DCs are defined, and detailed mathematical differentiation model are formulized and deducted. Lastly, several relations between quantity and lifecycle of different Agent category are proved. Simulation testing results demonstmte that DCs differentiation mechanism has theoretical significance and practical value on computer security besides decreasing false positive rate as well as achieving homeostasis for intrusion detection systems.

关 键 词：人工免疫 危险理论 树突状细胞 分化模型 信号融合 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]