改进的否定选择算法在入侵检测系统中的应用

摘要

NSA (Negative Select Algorithm) is one of important algorithms in artificial immune system, which is an imitation to developing procedure of immunize cell. The system gets mature detectors by it. The detector of traditional NSA has higher redundancy and lower coverage .Aim to the problem, the paper proposes a new method, which screens detectors twice, to reduce the duplication and redundancy. Then experiment demonstrate that it improves the correct detect rate by 10% and reduces the missing rate by 3%in IDS model. So the new algorithm is efficient.%人工免疫系统中的一个重要的核心算法就是否定选择算法（Negative Select Algorithm），该算法用来生成成熟检测器，它是对免疫细胞成熟过程的模拟。否定选择算法的检测器是随机生成的，因此检测器集合冗余度高、对非自体空间的覆盖率低。针对这一问题，本文提出一种改进算法，通过对检测器进行二次筛选来降低检测器的重复率和冗余度。实验证明，在基于免疫原理的入侵检测模型中改进算法可以将模型的正确检测率提高10%，漏检率降低3%，该算法是有效的。