Fulltext

Bandwidth Utilization Prediction in LAN Network Using Time Series Modeling

Shatha M. Hasan --- Mouayad A. Sahib, --- Ammar T. Namel

IRAQI JOURNAL OF COMPUTERS,COMMUNICATION AND CONTROL & SYSTEMS ENGINEERING المجلة العراقية لهندسة الحاسبات والاتصالات والسيطرة والنظم
ISSN: 18119212 Year: 2019 Volume: 19 Issue: 2 Pages: 78-89
Publisher: University of Technology الجامعة التكنولوجية

Abstract

monitoring the behavior of computer networks is essential forproblem identification and optimal management. Part of this behavior to bemonitored is the utilization of the network bandwidth. Several techniques areused to model and forecast network traffic such as time series models, moderndata mining techniques, soft computing approaches, and neural networks areused for network traffic analysis and prediction. Efficient bandwidth utilizationand optimization are very interesting research issues in effective networksbecause bandwidth is one of the most required and expensive Internetcomponents needed today. It is generally known that the higher the bandwidthavailable, the better the network performance, thus an essential aid for networkdesign and bandwidth wastage control and a need for trac models which cancapture the characteristics is necessary. In this paper, a time series predictionmodels were proposed for LAN office network bandwidth utilization. Theproposed prediction models are tested by using evaluation metrics used in timeseries such as MSE and performance evaluation plot. Testing results show thatthe proposed models can enhance the detection of bandwidth traffic and providean efficient tool for bandwidth utilization.

Keywords

Dynamic bandwidth allocation --- bandwidth consumption --- Time Series --- ARMA model --- AR model.