基于改进人工蜂群优化K- means的入侵检测模型

Intrusion Detection Model Based on Improved K-means of Artificial Bee Colony Optimization

在SD-WSN网络环境下,针对传统的入侵检测技术对不同种类的入侵者检测率低、检测速率慢等问题,提出了一种基于改进人工蜂群优化的K-means入侵检测模型。该模型首先筛选出SD-WSN网络所需的流量特征并对原始数据集进行预处理;其次,对传统的人工蜂群算法进行适应度函数的改进,加快模型的收敛速度,避免局部最优;再利用改进的人工蜂群算法对K-means算法的初始聚类中心进行优化选择,并对数据集进行训练形成入侵检测模型。利用该模型实现了对SD-WSN网络的控制器等网络实体的异常检测。实验结果表明,基于改进人工蜂群优化的K-means入侵检测模型较传统模型相比,检测率提高5%以上,误检率降低至4.5%以下,表现出良好的检测效果。

In the SD-WSN network environment,a traditional K-means intrusion detection model based on artificial bee colony optimization is proposed for traditional intrusion detection technology with low detection rate and poor detection speed for different types of intruders.Firstly,the traffic characteristics required by the wireless software defined sensor network is screened out and the original data set is preprocessed.Secondly,the traditional artificial bee colony algorithm is improved by the fitness function,which accelerates the convergence speed of the model and avoids local optimization.Then the improved artificial bee colony algorithm is used to optimize the initial clustering center of K-means algorithm,and the data set is trained to form an intrusion detection model.The model is used to detect the anomaly of network entities such as controllers of the SDWSN network.The experimental results show that the K-means intrusion detection model based on improved artificial bee colony optimization,compared with the traditional model,has a detection rate of more than 5%and a false detection rate of less than 4.5%,showing good detection results.