基于自监督学习的物联网终端网络行为监测方法

Behavior Monitoring Method of IoT Terminal Network Based on Self-Supervised Learning

常规的物联网终端网络行为监测方法以异常行为监测为主,入侵性较强的外部硬件设备很难满足大规模的监测需求,存在监测失误的问题。因此,设计基于自监督学习的物联网终端网络行为监测方法,基于自监督学习构建物联网终端网络行为监测模型,从物联网系统中获取海量的网络行为数据,将其作为输入数据训练自监督学习监测模型,从中提取有效的网络行为特征。监测物联网终端网络异常行为,根据物联网终端网络的实际情况,设置网络差异阈值,超过阈值后自动触发预警信号,确保物联网的安全使用。采用对比实验,验证该方法的监测性能更佳,能够应用于实际生活中。

The conventional network behavior monitoring method of the Internet of Things terminal mainly adopts abnormal behavior monitoring,and the highly invasive external hardware equipment is difficult to meet the large-scale monitoring needs,and there is the problem of monitoring error.Therefore,the IOT terminal network behavior monitoring method based on selfsupervised learning is designed.Based on self-supervised learning,the terminal network behavior monitoring model of the Internet of Things is constructed,obtaining massive network behavior data from the Internet of Things system,and training the self-supervised learning monitoring model as input data,from which effective network behavior features are extracted.Monitor the abnormal behavior of the terminal network of the Internet of Things,set the network difference threshold according to the actual situation of the Internet of Things terminal network,and automatically trigger the early warning signal after exceeding the threshold value to ensure the safe use of the Internet of Things.The comparative experiment is proved that the method is better and can be applied in real life.