免疫入侵检测中基于数据场的动态识别算法

Dynamic recognition algorithm based on data field in immune intrusion detection

将数据场理论引入到计算机免疫的研究中,设计了一种识别器的构造方法及其动态识别算法。抗体的培育是建立在不完全自体集的基础上,算法可以识别出未知自体,降低自免疫反应发生的概率,并通过动态识别算法完善抗体集,克服了现有的入侵检测系统对自体集要求较高的局限性,简化了克隆变异以及记忆机制的实现方法。实验表明:新的免疫动态识别方法使入侵检测系统具有更高的动态平衡性和自适应性。

A construction method of detector and its relevant dynamic recognition algorithm were put forward by introducing the data field theory to computer immunology. Antibodies are brought up based on self set. By recognizing the unknown self set, the algorithm can decrease the rate of self-immunity, and also improve the antibody set dynamically and overcome the limitations of traditional IDs that have high requirement for self set, thus simplify the way to implement cloning, mutation and memory. The results of experiments show that the new dynamic recognition algorithm makes IDs possess a higher self adaptability and dynamic equilibrium capability.