Applications Traffic Dynamic Early Warning Mechanism For The Electrical Power Integrated Data Network

With the continuous development of smart grid,it is more and more important to strengthen the application of power communication technology in the power grid.The electrical power integrated data network is the key communication infrastructure to ensure the safe and effective operation of the power grid.Now it has reached all levels and departments of power generation,operation and management.Therefore,it is necessary to establish a safe and reliable electrical power integrated data network.As an important guarantee mechanism for the quality and safety of electrical power integrated data network,the early warning for the traffic has a direct impact on the electrical power integrated data network's normal operation due to its accuracy and real-time performance.It is also the key to improve the operational efficiency of the power grid.With the rising of new applications such as video conferencing in the electrical power integrated data network,the fluctuation of traffic flow changes,making the traditional network traffic model no longer fit the existing network well.Therefore,it is urgent to propose a more targeted network traffic fitting model to meet the existing network application requirements.On the other hand,most of the existing detection algorithms cannot meet the requirements of macroscopic monitoring and anomaly detection for large-scale traffic data.In particular,the existing detection algorithms also increases the burden on network managers.Therefore,an adaptive and efficient anomaly traffic detection method is urgently needed.The purpose of this paper is to explore early warning technology of abnormal traffic in the current electrical power integrated data network,to improve the ability to detect and warn network traffic anomalies.Two aspects are researched in this article,including the network traffic prediction and dynamic threshold.First of all,analyze the application characteristics of the current network.Then divide the applications of the electrical power integrated data network into three categories of voice,video and data through the method of cluster analysis.Next,according to the different requirements of bandwidth and delay and the characteristics of suddenness and periodicity,learning the advantages of F-ARIMA model(fractional autoregressive integrated moving average model,F-ARIMA)and S-ARIMA(seasonal autoregressive integrated moving average model,S-ARIMA)model,to establish a Multi-Applications Comprehensive Traffic Prediction(MACTP)model to improve the prediction accuracy,and lay the foundation for the next step early warning.On the other hand,in order to improve the detection rate of traffic anomaly,sliding window method is used to calculate the dynamic baseline,and the initial value of the dynamic threshold is set considering the network bandwidth.Finally,the early warning is determined by comparing the traffic value given by the MACTP model with the upper and lower thresholds of the dynamic threshold,and the effectiveness and practicability of the early warning mechanism are verified through the simulation.