引入信息融合编码特征传递网络入侵检测仿真  

Network Intrusion Detection Simulation Based on Code Feature Transfer of Information Fusion

作　　者：龙文光[1] 

机构地区：[1]内江师范学院现代教育技术中心,四川内江641112

出　　处：《计算机仿真》2015年第7期315-318,共4页Computer Simulation

基　　金：校自然科学重点课题(14ZA03)

摘　　要：在网络信息交互和加密传输过程中,需要进行信息编码,由于编码特征的相异性,可能引入网络入侵信号,传统的网络入侵检测算法采用入侵信号包络线幅度检测,由于对幅度的阀值判断无法做到绝对精准,很容易引起的边界控制误差,形成信号相似性干扰,导致检测性能不好。提出一种信息融合编码特征传递的网络入侵检测算法,构建了低信噪比下的网络入侵信号模型,对信号进行降噪滤波处理,提高信号的纯度。通过自相关成分的独立快速分析与分离,得到入侵自相关检索编码,构建入侵检测目标函数,为使检测概率最优化,对检测目标函数求导,对入侵信号的幅度和频率进行参量估计,采用信息融合编码特征传递特征检测算法,实现对网络入侵检测模型改进。仿真结果表明,采用改进算法能有效提高网络入侵的检测性能,信号频谱分辨性较高,避免了边界误差,检测概率提高显著,性能优越,在网络与信息安全等领域具有较好的应用价值。An improved network intrusion detection algorithm is proposed based on information fusion coding characteristic transfer. The network intrusion signal model under low signal to noise ratio is constructed,the signal is processed by using a denoising filter,and the signal purity is improved. Through the independent rapid analysis and separation of self-correlation components,the self-correlation retrieval code of invasion is obtained,and intrusion detection target function is constructed. In order to optimize the detection probability,the detection target function is solved derivatively. The amplitude and frequency of intrusion signal are estimated,and the network intrusion detection model improvement is realized by using information fusion coding characteristic transfer feature detection algorithm. Simulation results show that the improved algorithm can effectively improve the performance of network intrusion detection,which has high signal spectrum resolution and can avoid the boundary error.

关 键 词：信息编码 网络入侵 检测算法 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]