中文摘要：

探讨了入侵检测系统的发展现状，研究生物免疫系统的特点和基本理论。在分析传统的人工免疫系统及LISYS系统的基础上，提出了一种改进的基于人工免疫的入侵检测系统。该系统在研究现有检测器生成算法的基础上，提出了位变异的初始检测器生成算法，对检测攻击的变异更为有效。该系统引入生物免疫学的协同刺激机制，并用LRU算法取代随机淘汰策略。实验结果证明这些方法能降低误报率，保证检测器的检测效率。

英文摘要：

The developing trend of intrusion detection technology is explained and the characteristic and basic theory of biological immune system is researched in this paper. An improved intrusion detection system based on artificial immune are proposed after Analyzing traditional artificial immune system and LISYS. Researching the generating algorithm of existing detectors, the improved system proposes initial detector＇s generating algorithm of bit variation. It makes more efficient for detect the variation of attacks. The co-stimulation mechanism of biological immunology is adopted and the strategy of random death is replaced by LRU algorithm in the system. The experiment shows that these methods can decrease the false positive rate and improve detection efficiency of detector.