Dynamic Traffic Anomaly Detection for Broadband Smart Grid Services in Software Defined Networks

The development of broadband smart grid services makes the network traffic increase rapidly. At the same time, it is also faced with the threat of various network attacks, which seriously affects the security of users. As a new type of network architecture, Software Defined Networks (SDN) offers new solutions to the management and optimization of network traffic. A dynamic detection algorithm for abnormal SDN traffic is proposed. The concepts of independence and deviation are defined, and the singularity, independence and deviation are used as the standard to calculate the three-probability p-values to judge the abnormal state of traffic, which is complicated in ensuring the algorithm time. Under the premise of high accuracy, the false alarm rate in the detection process is reduced.