基于多层特征基参数融合的网络入侵检测算法  

Network Intrusion Detection Algorithm Based on Multi-level and Parameters of the Fusion Feature

作　　者：李庆年[1] 

机构地区：[1]邕江大学,南宁530200

出　　处：《科技通报》2012年第8期69-71,共3页Bulletin of Science and Technology

摘　　要：传统的网络入侵检测算法,使用的都是同一个或者相似的特征基作为入侵检测的衡量标准,但是在多层网络中,不同层次的特征基参数存在差异,检测结果误差较大。为此提出了一种基于多层特征基参数融合的网络入侵检测算法。通过提取多层网络操作差异特征基参数,利用非线性回归方式对每层特征基参数进行差异补偿,按照自适应融合方式对多层差异特征基进行融合处理,以此作为检测的基础。实验表明,该算法提高了检测的准确率,取得了理想的效果。The traditional network intrusion detection algorithm,the use of all are the same or similar characteristics as the intrusion detection criterion,but in a multi-layered network,the different levels of the characteristics and differences between the parameters,the test result error is bigger.Therefore put forward based on the parameters of features of the integration of network intrusion detection algorithm.Through the extraction multilayer network operating difference characteristics and parameters by using nonlinear regression way for each layer,characteristic parameters of difference compensation,according to the adaptive way fusion of multilayer difference characteristics the fusion processing,as the basis of a test.Experiments show that the algorithm enhances the accuracy of inspection,to make the ideal effect.

关 键 词：入侵检测 多层网络 特征参数 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]