PERFORMANCE IMPROVEMENTS OF BIGBLUEBUTTON FOR DISTANCE TEACHING

Abstract

BigBlueButton (BBB) is a web conferencing system designed for online learning. It consists of a set of pre-configured open-source software tools to realize video conferencing functionality primarily for teaching purposes. Due to the COVID-19 pandemic, our university decided to roll out BBB for the university's educational activities in the first nationwide lock-down in early 2020. Based on our experiences in deploying, operating, and using BBB at our university for about 12 months, we present suggestions on how the services provided by BBB can be improved to meet the technical demands identified during online lecturing at our university. Our suggestions include the introduction of simulcast, improvements of encoding and muxing video feeds, and the 'Last -N' algorithm for video feed pagination. To demonstrate the benefits of the presented improvements, we experimentally evaluated most of them based on our own prototypical implementations.