关于交互式网络中异常数据准确检测仿真
|
摘要
对交互式网络中异常数据进行准确检测,能够有效提高交互式网络运行稳定性。交互式网络中异常数据的准确检测,首先需要对数据的小波量误差进行初步检测,对其在时间和空间上进行校正处理。传统方法考虑不同分布的异常攻击之间的差异,分析未知的交互式网络攻击下的异常数据行为,但忽略了对异常数据在时间与空间上的误差校正,导致异常检测精度偏低。提出基于最小二乘支持向量机的交互式网络中异常数据准确检测方法。通过构造一个基于异常数据测量值比例关系的F统计量,将其与异常数据约束残差统计量相结合,对稳态过程中出现的小波量误差进行初步检测,从时间冗余与空间冗余两个角度对异常数据在时间和空间上进行校正,完成交互式网络中异常数据的准确检测。实验结果表明,该方法可以准确检测出交互式网络中异常数据。
The accurate detection of abnormal data in interactive network can effectively improve the stability of interactive network. The traditional detection method often ignores the error correction of abnormal data in time and space,which results in the low accuracy. Therefore,this article presents an accurate detection method of abnormal data in interactive networks based on least squares support vector machine. By constructing an F statistic based on the proportional relation of measured values of abnormal data,we combine it with constrained residual statistics of abnormal data residual,and then detect wavelet error in the steady-state process initially. Finally,we correct abnormal data from the time redundancy and spatial redundancy,so as to complete accurate detection of abnormal data in interactive network. According to Simulation results,this method can detect abnormal data in interactive network accurately.
The accurate detection of abnormal data in interactive network can effectively improve the stability of interactive network. The traditional detection method often ignores the error correction of abnormal data in time and space,which results in the low accuracy. Therefore,this article presents an accurate detection method of abnormal data in interactive networks based on least squares support vector machine. By constructing an F statistic based on the proportional relation of measured values of abnormal data,we combine it with constrained residual statistics of abnormal data residual,and then detect wavelet error in the steady-state process initially. Finally,we correct abnormal data from the time redundancy and spatial redundancy,so as to complete accurate detection of abnormal data in interactive network. According to Simulation results,this method can detect abnormal data in interactive network accurately.
引文
[1]王洪乐,等.基于数控机床综合误差补偿方法[J].西安交通大学学报,2017,51(6):138-146.
[2]栾元重,等.桥梁变形监测数据小波去噪与Kalman滤波研究[J].大地测量与地球动力学,2015,35(6):1041-1045.
[3]李东风,周霞.网络数据库中海量数据压缩传递方法研究仿真[J].计算机仿真,2016,33(5):196-199.
[4]刘国奇,等.基于小波神经网络的风机故障诊断[J].小型微型计算机系统,2015,36(7):1504-1508.
[5]刘科研,等.基于多源多时空信息的配电网SCADA系统电压数据质量检测与评估方法[J].电网技术,2015,39(11):3169-3175.
[6]邹士迁,刘鑫,陈卫华.基于实测光电跟踪仪误差的分布检验及应用[J].火力与指挥控制,2017,42(3):174-178.
[7]赵新杰,刘渊,孙剑.基于迁移学习和D-S理论的网络异常检测[J].计算机应用研究,2016,33(4):1137-1140.
[8]李宇翀,等.RMPCM:一种基于健壮多元概率校准模型的全网络异常检测方法[J].通信学报,2015,36(11):201-212.
[9]姜红红,等.基于大数据的电力信息网络流量异常检测机制[J].电信科学,2017,33(3):134-141.
[10]苏维龙.交互式CAI课件中的分段路径规划[J].电子设计工程,2016,10(24):58-60.
[2]栾元重,等.桥梁变形监测数据小波去噪与Kalman滤波研究[J].大地测量与地球动力学,2015,35(6):1041-1045.
[3]李东风,周霞.网络数据库中海量数据压缩传递方法研究仿真[J].计算机仿真,2016,33(5):196-199.
[4]刘国奇,等.基于小波神经网络的风机故障诊断[J].小型微型计算机系统,2015,36(7):1504-1508.
[5]刘科研,等.基于多源多时空信息的配电网SCADA系统电压数据质量检测与评估方法[J].电网技术,2015,39(11):3169-3175.
[6]邹士迁,刘鑫,陈卫华.基于实测光电跟踪仪误差的分布检验及应用[J].火力与指挥控制,2017,42(3):174-178.
[7]赵新杰,刘渊,孙剑.基于迁移学习和D-S理论的网络异常检测[J].计算机应用研究,2016,33(4):1137-1140.
[8]李宇翀,等.RMPCM:一种基于健壮多元概率校准模型的全网络异常检测方法[J].通信学报,2015,36(11):201-212.
[9]姜红红,等.基于大数据的电力信息网络流量异常检测机制[J].电信科学,2017,33(3):134-141.
[10]苏维龙.交互式CAI课件中的分段路径规划[J].电子设计工程,2016,10(24):58-60.