Internet measurement for enhanced performance in the COVID-19 era.

Abstract

This project entails carrying out internet measurements to the online learning resources used by makerere students, utilizing both on campus and off campus measurements with the main aim of identifying any limitations on the network that affect connectivity. This project was done using RIPE Atlas Software which was installed at CEDAT and a PerfSONAR toolkit which was installed at COCIS, where the servers to these resources are hosted and managed by DICTS. These toolkits were used to test the quality of connectivity using network parameters like throughput, packet loss, latency and topology (traceroute). The key results from this project were identification of usage patterns from the students and a congestion point on the network with an instance of very high latency up to 30ms. The team concluded that this congestion point, using the network parameter values from both toolkits, was the contributing to instances of poor connectivity when many students are trying to access the online learning resources. It was also discovered, from both toolkits, that the congestion point on the network was an Internet Service Provider (ISP) termination point where all incoming requests pass from on and off campus network. The team recommends adding more ISP termination points or increasing the bandwidth of that part of the network.