A Fast Method for Identifying Anomalous Traffic in a Ubiquitous Power IoT Access Layer Network

Conventional IoT access layer network abnormal traffic rapid identification nodes are mostly deployed in a one-way structure, and the identification coverage is limited, resulting in an increase in the final error recognition rate. Therefore, this paper proposes a design and identification comparison of ubiquitous IoT access layer network abnormal traffic rapid identification methods. According to the current identification requirements, firstly extract the characteristics of abnormal traffic, and use a multi-level approach to break the limit of identification coverage, so as to realize the deployment of multi-level identification nodes. On this basis, design an adaptive fast identification model for abnormal traffic in the access layer network of the power Internet of Things, and finally achieve fast identification processing by using dynamic and continuous monitoring. The test results show that compared with the SDN abnormal traffic detection system with different mathematical models for monitoring and identifying abnormal network traffic methods and deep learning hybrid models, the error rate of the rapid identification method of abnormal traffic in the access layer network of the ubiquitous power Internet of Things designed this time is relatively low, which indicates that the rapid identification mode of abnormal network traffic designed this time is more flexible, efficient, safe and targeted. In a complex background environment, the identification effect of traffic is significantly improved.