分层异构网络链路中断故障检测仿真
|
摘要
为了保证分层异构网络信息传输的正常运行,提高无线通信系统的稳定性,需要进行网络链路中断故障检测,及时维护并修复网络链路。针对当前分层异构网络链路中断故障检测方法中存在的误警率较高,检出率较低,检测效率较低等问题,提出基于断点数据特征的分层异构网络链路中断故障检测方法。分析网络链路及各节点的关系,计算节点信息发射功率与信噪比,据此求算网络链路中断概率,确定检测范围;对网络链路断点数据进行样本采集,结合时频分析法提取样本数据的特征,通过聚类分析划分断点数据特征类型;完成断点数据特征提取与分类后,结合线性变换与KNN算法对分层异构网络链路数据进行分类,划分正常数据与异常数据,完成链路中断故障检测。实验结果表明,所提方法的故障检测误警率低于实验对比方法,故障检出率较高,且检测过程消耗时间更短。
Due to high false alarm rate,low detection rate,and poor detection efficiency of current fault detection method of link interruption of hierarchical heterogeneous network,a new fault detection method based on breakpoint data feature is proposed in this research. Firstly,our research analyzed relationship between network link and each node,and calculated transmit power and signal to noise ratio(SNR) of node information. According to that,the research calculated interruption probability of network link and confirmed detection range. Then,the research collected sample of the breakpoint data and extracted feature of the sample data integrated with time-frequency analysis method to divide type of the breakpoint data feature via cluster analysis. After extraction and classification of the breakpoint data feature,the research classified data of the hierarchical heterogeneous network link integrated with linear transformation and KNN algorithm and divided normal data and abnormal data. Thus,we completed the fault detection. Simulation results show that false alarm rate of the method is much lower and the detection rate is higher. Meanwhile,detection time consumes much less time.
Due to high false alarm rate,low detection rate,and poor detection efficiency of current fault detection method of link interruption of hierarchical heterogeneous network,a new fault detection method based on breakpoint data feature is proposed in this research. Firstly,our research analyzed relationship between network link and each node,and calculated transmit power and signal to noise ratio(SNR) of node information. According to that,the research calculated interruption probability of network link and confirmed detection range. Then,the research collected sample of the breakpoint data and extracted feature of the sample data integrated with time-frequency analysis method to divide type of the breakpoint data feature via cluster analysis. After extraction and classification of the breakpoint data feature,the research classified data of the hierarchical heterogeneous network link integrated with linear transformation and KNN algorithm and divided normal data and abnormal data. Thus,we completed the fault detection. Simulation results show that false alarm rate of the method is much lower and the detection rate is higher. Meanwhile,detection time consumes much less time.
引文
[1] 宋康,韦磊,冀保峰,等.基于干扰消除的异构蜂窝网络中断分析[J].电子与信息学报,2016,38(2):255-261.
[2] 薛华威,王宝生,邓文平,等.基于SDN架构的网络故障检测与修复系统[J].计算机工程,2017,43(11):40-44.
[3] 马忠贵,刘立宇,闫文博,等.基于泊松簇过程的三层异构蜂窝网络部署模型[J].工程科学学报,2017,39(2):309-316.
[4] 吕莉.软件分层结构故障优化检测仿真研究[J].计算机仿真,2017,34(10):371-374.
[5] 赵亮,王旭东,裴强强,等.G7公路甘肃段爆破施工对土遗址震动影响的检测与分析[J].地震工程学报,2017,39(2):343-347.
[6] 邓丰,曾祥君,马士聪,等.基于分布式行波检测的广域网络故障定位方法[J].电网技术,2017,41(4):1300-1307.
[7] 王军晓,齐恒,李克秋,等.面向数据中心网络的链路故障实时检测即服务[J].计算机研究与发展,2018,55(4):704-716.
[8] 李靖,彭海云.基于分布式算法执行替代链的WSN故障节点检测和替代方法[J].计算机应用研究,2016,33(2):557-561.
[9] 逯程徐廷学,王虹.基于范数正则化的塑性回声状态网络故障诊断模型[J].北京航空航天大学学报,2018,44(3):535-541.
[10] 夏晓峰,何常胜.LSM结合邻居干扰抵抗模型的传感器网络节点故障检测[J].湘潭大学自然科学学报,2016,38(1):101-105.
[2] 薛华威,王宝生,邓文平,等.基于SDN架构的网络故障检测与修复系统[J].计算机工程,2017,43(11):40-44.
[3] 马忠贵,刘立宇,闫文博,等.基于泊松簇过程的三层异构蜂窝网络部署模型[J].工程科学学报,2017,39(2):309-316.
[4] 吕莉.软件分层结构故障优化检测仿真研究[J].计算机仿真,2017,34(10):371-374.
[5] 赵亮,王旭东,裴强强,等.G7公路甘肃段爆破施工对土遗址震动影响的检测与分析[J].地震工程学报,2017,39(2):343-347.
[6] 邓丰,曾祥君,马士聪,等.基于分布式行波检测的广域网络故障定位方法[J].电网技术,2017,41(4):1300-1307.
[7] 王军晓,齐恒,李克秋,等.面向数据中心网络的链路故障实时检测即服务[J].计算机研究与发展,2018,55(4):704-716.
[8] 李靖,彭海云.基于分布式算法执行替代链的WSN故障节点检测和替代方法[J].计算机应用研究,2016,33(2):557-561.
[9] 逯程徐廷学,王虹.基于范数正则化的塑性回声状态网络故障诊断模型[J].北京航空航天大学学报,2018,44(3):535-541.
[10] 夏晓峰,何常胜.LSM结合邻居干扰抵抗模型的传感器网络节点故障检测[J].湘潭大学自然科学学报,2016,38(1):101-105.