{0基于免疫状态转换的网络态势觉察模型的研究}0{.doc

{0基于免疫状态转换的网络态势觉察模型的研究}0{Network state Suspicion model Research Ground on Ground On immune state immune state transformed0}. {0引言}0{Introduction0}. {0在网络中，随着环境、时间、要求的改变，网络安全检测要求能够更加高效、自适应、可扩展，以确保整个网络的完整性、机密性和有效性。}0{In etwork ， with circumstance , hour, and required hanged ， etwork Security detection request can all the more highly active , self adapting , extendible, to insure entire network’s perfection, confidentiality and usability. 0}. {0网络安全检测系统基于入侵检测系统的基础主要使用异常检测和误用检测两种技术。}0{Ground n invasion detection system detection system’s network security detection system mostly use anomaly detection and misuse detection two technology. 0}. {0而这些与生物免疫有非常相似的地方，因此人工免疫用于网络安全检测已经取得一定的成果。}0{Whereas these have much similar place with reature Immunization, so artificial immunity design to Network Security detection already taking definite payoff. 0}. {0最初就是Forrest教授提出的分布式人工免疫系统模型ARTIS[55][56][57]，以及后面Ballet[58]教授提出的基于免疫原理的分布式multi-agent计算机免疫系统和Kim[59][60][61]教授提出的基于动态克隆选择的网络入侵检测模型等等。}0{At irst distributing type artificial immunity system model ARTIS[55][56][57] by professor Forrest, and also distributing type multi -agent computer immune system Ground On immunogenic rational by professor Ballet [58] and Network Invasion detection model ground on dynamic state colonel selection by Professor Kim [59][60][61] and so on .0}. {0但这些研究成果并不完善，存在部分缺陷。}0{Yet these returns none so soundnessresence part of defect. 0}. {0如对网络安全检测系统中的自体/非自体采用为静态描述，良好的适应性，不能满足真实网络环境下的需求，使得系统误报率、漏报率较高。}0{For instance check the net Security detection systems auto/no check adopt for static tate description, flexibility favorable, insatiability wants and needs of verity network environment, so system misstatement rate , missing rate higher .0}. {0另一方面免疫检测细胞（检测器）在其进化淘汰过程中缺乏动态缓冲机制，一些对入侵检测有重要作用的检测器仅仅由于在某一时段原来与之匹配的非自体转换为自体造成该检测器耐受失败被抛弃。}0{On the other and immunes detection corpuscle(detector) in that evolution weed course thong lack dynamic buffering Mechanism, some to invasion detection ponder ability want funct