TrioPen: A novel model to prioritize responsive flows enabling enhanced multimedia communication on the Internet

The use of social media platforms around the world has exponentially increased Internet traffic. Such platforms enable multimedia communication including textual, audio, and video streaming. This adds enormous amount of data to the network, which adversely impacts the network throughput. A responsive source employs a reliable adaptive policy for data transmission among nodes, which is exploited by unresponsive source nodes to their favour that leads towards unfair queue allocation. The existing literature provides several queue management models, however, these models are incapable of penalizing unresponsive flows to avoid unfair queue occupation. To resolve this issue, we propose a novel Triple penalty-enabled prioritized queue management (TrioPen) model, which imposes a three-tier penalty to prioritize responsive flows. TrioPen splits the physical queue into two virtual queues, where a predefined threshold is maintained on the first virtual queue, as a tier-1 penalty, to restrict unresponsive flows. Moreover, TrioPen introduces a novel scheduling mechanism for transition of packets among the virtual queues, where a skip factor works as the tier-2 penalty and increases queueing delay for unresponsive flows in comparison with responsive flows. Additionally, another predefined threshold in the second virtual queue, serving as tier-3 penalty, further restricts unresponsive flows, thereby, ensuring fair queue allocation among responsive and unresponsive flows. Simulation results demonstrate that TrioPen reduces packet loss rate to 1.91 %, end-to-end delay to 0.04 ms, jitter to 0.05 ms, and improves throughput to 98.05 % for responsive flows, which is significantly better in comparison with eminent queue management models.