大数据环境下的网络主动入侵检测方法研究

发布时间：2018-03-06 11:35

本文选题：大数据　切入点：网络入侵　出处：《科技通报》2015年08期 　论文类型：期刊论文

【摘要】：由于传统网络入侵检测方法检测率低及不能进行在线检测,无法有效实现网络主动入侵检测,提出一种基于马氏距离K均值的大数据环境下网络主动入侵检测方法,分析了马氏距离的评价准则,依据新样本与原样本之间的马氏距离确定是否需开展新的聚类,输出与全部数据样本相应的攻击类别。通过一个由一定数量的支持向量决定的超平对数据进行分类,当SVM分离方法受到约束时,利用核函数将输入数据映射至高维特征空间,采用高斯径向基函数对最小二乘支持向量机分类模型进行建立。通过粒子群优化算法对最小二乘支持向量机参数进行选择。利用种群中个体之间的协作以及信息交换获取最佳方案。仿真实验结果表明,所提方法有很高的的检测效率及精度。
[Abstract]:Due to the low detection rate of traditional network intrusion detection methods and the inability to carry out on-line detection, the active network intrusion detection method based on Mahalanobis distance K-means in big data environment is proposed. The evaluation criteria of Markov distance are analyzed. According to the Markov distance between the new sample and the original sample, whether a new clustering is needed or not is determined. When the SVM separation method is constrained, the input data is mapped to the high dimensional feature space by the kernel function when the data is classified according to the attack class corresponding to all the data samples, and a certain number of support vectors are used to classify the data, and when the SVM separation method is constrained, the kernel function is used to map the input data to the high dimensional feature space. The classification model of least squares support vector machine is established by Gao Si radial basis function. The parameters of least squares support vector machine are selected by particle swarm optimization algorithm. The cooperation and information among individuals in the population are used. The simulation results show that, The proposed method has high detection efficiency and accuracy.
【作者单位】：湖北工程学院计算机与信息科学学院;
【基金】：国家自然科学基金(61370092) 湖北省自然科学基金(No.2013CFC005) 湖北省高等学校优秀中青年科技创新团队计划(T201410)
【分类号】：TP393.08

【参考文献】