基于Poisson过程和Rough包含的计算免疫模型  

An Computing Immune Model Based on Poisson Procedure and Rough Inclusion

作　　者：励晓健[1] 黄勇[1] 黄厚宽[1] 

机构地区：[1]北方交通大学计算机科学技术系,北京100044

出　　处：《计算机学报》2003年第1期71-76,共6页Chinese Journal of Computers

摘　　要：将随机过程引入计算机免疫的研究中 ,提出了一种新的计算免疫模型 ;分析了访问端口的参数 ,引入Pois son随机过程 ,从而将串比较转化为数值计算以解决比较的效率问题 .经过随机过程的计算 ,可以较为精确地确定当前的异常访问模式可能属于哪种类型的入侵模式 ,从而为下一步入侵模式的匹配大大缩小了搜索范围 .According to observing port visit request of server carefully and discovering the Possion procedure of port visit request, authors introduce Possion stochastic procedure theory and propose a immune model based on Possion procedure. The introduction of Possion procedure for immune system convert string comparison into mathematic computation.The advantages include three aspects.First,This model saves much storage because saving many normal visit strings is not necessary in detector generating region.Second,it improve the effiency of comparison.Third,it enhances adaptivity and flexibility of immune system .It is unnecessary for immune system to change many normal visit strings when visit request state is changed.And the immune system only modulates its parameters viz. λ of every port to adapt this change. Also to resolve string match problem, authors adopt some methods of rough set. The experiment results demenstrate that this model is applicable.Because there are many stochastic procedures in our world, the model proposed in this paper is applicable to many domains when it is somewhat modified,such as classification,machine fault detection,finance tendency analysis,model recognition and so on.

关 键 词：POISSON过程 Rough包含 计算免疫模型 计算机安全 计算机免疫系统 粗糙集合 入侵检测 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]