Bandwidth Performance Analysis and Shaping Algorithm on Metro-E Campus Network

• Published in: Colloquium in Information Science and Technology, CIST
• Main Author: Shamsuddin A.D.S.; Kassim M.; Abdullah N.P.
• Format: Conference paper
• Language: English
• Published: Institute of Electrical and Electronics Engineers Inc. 2023
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85185403318&doi=10.1109%2fCiSt56084.2023.10409961&partnerID=40&md5=7461082caf11f921d4659e6d79868a1f

This paper presents the analysis of the traffic of Metro Ethernet (Metro-E) Network (MEN) in the campus network. MEN's technology has given financial manageability to all clients and the expense of utilizing this service has increased from time to time but lacks Quality of Service (QoS) on the accessibility of the clients. This research aims to analyze five (5) campus bandwidth performances on the Metro-E network. Distribution modeling was analyzed based on the data captured and policing algorithm was deployed to identify QoS performance. Data on Campus Network bandwidth traffic was collected using Metro E monitoring tools on the CACTI domain then data are fitted to the best distribution models. Five best-fitted distributions are identified based on maximum likelihood estimator (MLE) value. Next, a shaping algorithm is designed to control the bandwidth used based on a 10% upper bandwidth and fixed threshold. The result presents that the five best traffic distributions are identified which are normal, stable, exponential, logistic, and lastly T-Location Scale. Among the 5 distributions model, it is identified that 'Stable Fits' for Campus A presents the best MLE value of -121642.QoS using a shaping algorithm has been successfully done where 10% upper bandwidth has been derived from the collected traffic data. This research is significant for QoS on bandwidth management on Metro Ethernet Campus Networks where bandwidth can be controlled to help the campus to reduce the cost of underutilized bandwidth at nonpeak hours. © 2023 IEEE.