现有的基于免疫机制的NIDS模型有着诸多缺陷。针对这些缺陷,同时借鉴关于免疫识别的最新理论(危险模式理论),本文提出了一个新型的NIDS模型。其中,改变了自体库的角色,以减轻检测代理的负担;引入了基于异常信号的免疫识别,降低了虚警率,同时增强了系统的适应性;提出了新的亲和力算法,通过“关键基因片段控制串”提高了识别的效率和准确率。 The existing NIDS model based on immune mechanism has many defects. In response to these shortcomings, at the same time drawing on the latest theory of immune recognition (dangerous mode theory), this paper presents a new NIDS model. Among them, the role of self-library was changed to reduce the burden on testing agents; the immune recognition based on abnormal signals was introduced to reduce the false alarm rate and enhance the adaptability of the system; a new affinity algorithm was proposed, Gene fragment control strings improve the efficiency and accuracy of identification.