基于人工蜂群算法的分布式入侵攻击检测系统

针对网络入侵攻击检测系统检测准确率与计算效率较低的问题,提出一种基于人工蜂群算法的分布式入侵攻击检测系统。将训练集划分为若干的子集,使用特征选择方法提取特征集中类内相关性高、类外相关性低的特征;对人工蜂群算法进行修改,通过引入全局搜索能力强的算法提高人工蜂群算法的性能;根据优化的特征子集与规则集对网络入侵攻击行为进行分类处理。基于网络入侵数据集的实验结果表明,该系统实现了较高的检测性能和计算效率。

DISTRIBUTED INTRUSION DETECTION SYSTEM OF NETWORKS BASED ON ARTIFICIAL BEE COLONY ALGORITHM

Aiming at the problems of low detection accuracy and low computational efficiency of intrusion detection system of networks, we presented a distributed intrusion detection system of networks based on modified artificial bee colony. The training set was divided into several subsets, and we used feature selection methods to abstract the features of high inner-class correlation and low intra-class correlation. Then the artificial bee colony algorithm was modified, and the algorithm with strong global search capability was introduced to improve the performance of artificial bee colony algorithm. The intrusion attack behaviors of networks were classified based on the optimized feature sets and rule sets. Results of the experiment based on the datasets of networks intrusion indicated that the proposed system realizes a good detection performance and computational efficiency.