Network Congestion Analysis using ARIMA Model and Bandwidth Cost Prediction

Abstract

Analysis and prediction of telecom network traffic as well as computer network traffic has a good scope in extensive set of areas and possess significant number of studies. Congestion in the network is one of the major causes of performance degradation which effect the whole communication network. Network Congestion analysis and its prediction is a better approach to ensure safe and authentic network communication. Different techniques have been proposed and experimented for analysing the traffic in network. This paper applies Time Series analysis and used Arima model to analysis the congestion in the network and based on the analysis the model predicts the cost based on the required bandwidth which is the need of different telecom company. By using Arima model the system proposed weather the congestion in the network is sessional or any trend in the network congestion which helps us to find when the congestion in the network is high and when it is low so that the bandwidth will be provide those telecom company so that they  increase the performance of the network.