Intelligent Bandwidth Planner Enhancing Learning Technologies in Constrained Distance Learning Environments – a Pandemic Response

Bandwidth resource in open distance electronic learning platform is scarce when services performed by many users contend for bandwidth, causing congestion in the network. The challenge has increased tremendously since almost all academic institutions perform activities online during the pandemic. This paper investigates the inherent workload constraints among open-distance electronic learning (ODeL) services competing for scarce resources and intends to forecast future bandwidth demands to prevent online class disruptions. A predictive framework of bandwidth management, which integrates a sustainable hidden Markov model (HMM) and a normalization policy, coupled with SolarWinds technology for prior network data feeder, is developed. A sustainable HMM $\alpha$ emerges from three HMM candidates based on test priorities on bandwidth demands. Compared to four popular methods, detailed experiments on the proposed model revealed performance analysis of error metrics below 0.5 at peak and off-peak periods. The emerged HMM $\alpha$ reliably predicted bandwidth capacities required to sustain the competing ODeL services with an accuracy of 94%.