QARIMA: A new approach to prediction in queue theory

One of the important issues in computer networks is Active Queue Management (AQM) that increases the performance of the network. Autoregressive Integrated Moving Average (ARIMA) as an Active Queue Management can improve methods such as congestion control and flow control by predicting the state of the queue in the networks. In current complications of queue theory in computer networks, due to the lack of linear constant trend and other issues such as packets bursting and non-periodic fluctuations of queue length, the present methods of prediction are being challenged. In this paper, a new anticipation ploy is proposed, which improved the prototype of ARIMA by considering available problems and requirements in computer networks. The subscribed algorithm that is called Queue-based ARIMA (QARIMA), can present predictions which are closer to true data, by uplift input data models.