基于二进制QPSO约简算法的入侵检测模型

Intrusion-Detection Model based on Binary QPSO Reduction Algorithm

为了克服入侵检测系统存在着在先验知识较少情况下推广能力差的问题,提出了基于粗糙集理论的入侵检测方法。将基于断点重要性的算法应用于本模型的离散化中,将二进制量子粒子群算法(QPSO)应用于属性约简中加以仿真,并与基于辨识矩阵和遗传算法的属性约简方法进行比较,结果表明,基于二进制量子粒子群算法的约简方法约简时间短、检测率高,优于辨识矩阵和遗传算法。

In order to overcome poor generalizing ability of current intrusion detection system in case of less prior knowledge, the intrusion detection method based on rough set theory is proposed. The algorithm based on breakpoint importance is applied in the discretization of the model, and the binary QPSO（quantum particle swarm optimization） in the attribute reduction, its simulation and comparision with attribute reduction methods based on discernibility matrix and genetic algorithm indicate that the reduction method based on the binary quantum particle swarm algorithm has the advantages of shorter reduction time and higher detection rate, and is superior to the identification matrix and genetic algorithm.