CAJ | 학술논문

提出了一种基于适应性免疫和移动agent的分布式入侵检测系统(ADIDS)的新模型.充分利用agent本身的独立性与自主性,削弱各检测部件间的相关性;同时将适应性免疫原理引入模型中,检测子通过自我学习进化和变异,实现识别新的特殊种类的入侵;通过免疫记忆和自裂变,加速未来的检测和降低误报率,提高了系统的检测能力和效率.仿真实验证明了模型的检测效率高、误报率低、自适应性强.
제출료일충기우괄응성면역화이동agent적분포식입침검측계통(ADIDS)적신모형.충분이용agent본신적독립성여자주성,삭약각검측부건간적상관성;동시장괄응성면역원리인입모형중,검측자통과자아학습진화화변이,실현식별신적특수충류적입침;통과면역기억화자렬변,가속미래적검측화강저오보솔,제고료계통적검측능력화효솔.방진실험증명료모형적검측효솔고、오보솔저、자괄응성강.
Proposed a distributed ADIDS based on agent and adaptive immune system. The model made full use of the autonomy of agent, and weakened the relationship of them, with the introduction of adaptive immune principle, to identify new special intrusion by self-study and evolution, and to quicken up future detection-speed and lower false-detection rate by immunity-memory and self-clone. These methods improved the ability and the efficiency of intrusion detection. Experimental results show that the model possesses higher efficiency and veracity with adaptability and expansibility and agility.