基于孤立森林的移动终端网络数据异常检测

Data anomaly detection of Mobile terminal network based on isolation forest

为了提高移动终端网络数据安全,提出基于孤立森林的移动终端网络数据异常检测方法。将移动终端数据流视作网络流量,根据信息增益原理完成数据片段划分,获取数据片段信息特征,并采用归一化处理验证数据特征可靠性。使用孤立森林算法中异常分数算法对获取到的数据特征进行数据异常检测,并设定相应的判定条件完成检测过程。构建实验环节,通过模拟移动终端数据传输的形式,验证文中方法与传统方法的使用效果,文中方法的检测率与误差率的实验结果均优于传统方法。文中方法对于动态数据的检测能力较佳,对网络数据具有较高的控制能力。

In order to improve the security of mobile terminal network data,an outlier detection method of mobile terminal network data based on isolation forest is proposed.The mobile terminal data stream is regarded as the network traffic,and the data segment is divided according to the information gain principle to obtain the information characteristics of the data segment,and the data feature reliability is verified by normalization processing.The outlier fraction algorithm of the isolation forest algorithm is used to detect the anomaly of the obtained data features,and the corresponding judgment conditions are set to complete the detection process.Through the simulation of mobile terminal data transmission,the effect of this method and the traditional method is verified.The detection rate and error rate of this method are better than the traditional method.In conclusion,this method has better detection ability for dynamic data and higher control ability for network data.