The Prediction of Demands for Bandwidth in Computer Network Through Fuzzy Time Series Song-Chissom Method

Abstract

The increasing of utilization and development of internet network infrastructure in higher educational institutions influences highly on the increasing of demands for bandwidth in computer network, hence it is necessary to have an efficient, reliable, yet economical service system. It can occur by conducting an appropriate and good system planning. Prediction of demands for bandwidth is one of supporting factors to plan an efficient and reliable internet servive system. This study implemented fuzzy time series Song-Chissom method adjusted the academic calendar. The forecasting system of fuzzy time series captures the pattern of prior (past) data to project the following (future) data. From the testing results obtained by using Song-Chissom method, it was found the mean error of 7.669% for course condition. From the result of prediction of demands for bandwidth, it is proposed 16 Mbps for course condition in State Polytechnic of Sriwijaya.

Key Words: Prediction/Forecasting, Bandwidth, Fuzzy time series

The authors decided to use the pseudonym and remove their real names in full article.