基于改进蚁群算法与遗传算法组合的网络入侵检测

为提高网络入侵检测的检测效果,提出一种基于改进蚁群算法与遗传算法组合的网络入侵检测方法.该方法采用遗传算法(genetic algorithm,GA)对网络入侵的特征集进行快速选取,为后续特征提取打下基础;对传统蚁群算法(ant colony optimization,ACO)的节点选择策略和信息素更新策略进行改进,提出一种改进的蚁群算法,提高对最优特征的选择效果,采用改进的蚁群算法对特征进一步选择;采用支持向量机(support vector machine,SVM)统计机器学习方法建立各类网络入侵的检测分类器.仿真实验结果表明,新的网络入侵检测方法综合GA和改进蚁群算法的优势,能够获得更好的入侵特征,从检测正确率、误报率和漏报率3个方面综合比较,新的网络入侵检测方法具有更好的网络入侵检测效果,且提高了检测速率.

Network intrusion detection method based on combination of improved ant colony optimization and genetic algorithm

To improve the detection effect of network intrusion detection, a network intrusion detection method based on improved ant colony algorithm in combination with Genetic Algorithm (GA) was proposed. Firstly, GA was used for rapid selection of network intrusion feature set for the following feature selection. Then an improved ant colony algorithm was proposed by the improvement of node selection and pheromone updating strategy to improve the detecting effect of feature, the improved ant colony algorithm was adopted for further feature selection. Finally, the method adopted support vector machine (SVM) statistical machine learning method to establish a network intrusion detection classifier. Simulation results show that the new network intrusion detection method merges the advantages of GA and improved ant colony algorithm, which can get better feature detection results. The comparison in terms of detection accuracy, false alarm rate and missing report rate shows that the new network intrusion detection method can get better network intrusion detection results, and the detection rate was also improved.