网络信息传输异常数据检测仿真研究

Research on the Simulation of Abnormal Data Detection in Network Information Transmission

对网络信息传输的异常数据进行检测,能够有效保证网络安全,防止网络异常瘫痪。对异常数据的检测,需要确定实现信息分类信息熵,并对数据属性的增益进行计算,完成信息传输异常数据的检测。传统方法对观测数据中网络信息传输异常数据进行识别,利用响应函数对数据进行投影分析,但忽略了获取数据属性增益。提出基于信息熵的网络信息传输异常数据检测分析模型建立方法,通过信息增益的方法作为属性选择的标准,确定生成决策树每个节点所需要的合适属性,并利用向量空间中正例集和反例集确定实现分类所需要的信息熵,并对数据属性的增益进行计算,实现数据分类,提高网络信息传输异常数据提取的效果。确定数据之间的异常程度,提出基于信息熵的网络信息传输异常数据检测分析模型建立方法。实验结果表明,所提方法能够准确检测出网络信息传输异常数据,且计算耗时较少。

To detect abnormal data in network information transmission can effectively guarantee the security of network. The detection for abnormal data needs to determine the information entropy which realizes information classification, and calculate the gain of data attribute, so as to complete the detection of abnormal data in information transmission. The traditional method ignores the gain of data attribute. In this paper, we focused on a method for building the detection and analysis model of abnormal data in network information transmission based on information entropy. The method of information gain was used as the standard of attribute selection to determine the appropriate attribute required by each node in generating decision tree. Then, positive example set and negative example set in vector space were used to information entropy needed to achieve the classification. Moreover, the gain of data attribute was calculated to realize data classification and improve effect of abnormal data extraction in network information transmission. Finally, the anomaly degree between data was determined. According to simulation results, the proposed method can accurately detect the network information transmission abnormal data and needs the less time computation.