ABSTRACT
Bandwidth utilization in a campus network if not properly managed,may leads to the congestion of the network bandwidth. This research work presents bandwidth control system implemented at the access layer of a campus network which is aimed at reducing the processing time at the core of the network. Ibrahim BadamasiBabangidaUniversity Lapai (IBBUL) campus network was monitored and data were collected on daily basis between 12am to 11pm over a period of 60 days (from 1st September, 2014 to 30th October, 2014). Analyses of these data were carried out which include: (i) peer-to-peer (p2p) activities, (ii) user traffic analysis and,(iii) web analysis. The analyses revealed the highest and lowest p2p activities of the users across the monitored nodes, as well as the time-of-day they occurred. It also revealed the nodes with the highest traffic and the time-of-day it usually occurred. Based on these results, IBBUL network scenario was modelled and simulated using Graphical Network Simulator 3 (GNS3). A star network topology with a Juniper EX4200 core was modelled. Juniper EX2200 series switches were modelled as the access layers. A number of host computers with various specifications were also attached to these switches and all the devices were configured appropriately. Traffic shaping technique was employed to implement bandwidth regulation on the access layer switches. The results obtained were validated by comparing it with the monitored data collected. From the results it was noticed that the developed model performed better by reducing the average processing time of the core to 6.00ms when bandwidth management was implemented at the access layer as compared to the average processing time of 9.950ms when bandwidth management was implemented at the core. This represents about 40% reduction in processing time. The average rate of download and upload speeds of the network users increased by 203.76Kb/s and 110.07Kb/s respectively. This represents about 95% improvement in the download and upload speeds.