Edge-Hosted ABR Control: 5G MEC Trials and QoE Gains
DOI:
https://doi.org/10.32996/jcsts.2025.4.1.81Keywords:
Multi-Access Edge Computing, Adaptive Bitrate Control, 5G Networks, Video Streaming Optimization, Quality of Experience, Multi-Armed Bandit AlgorithmsAbstract
Mobile video streaming technology is subject to serious challenges based on inherent latency constraints of conventional cloud-hosted adaptive bitrate control systems, sacrificing responsiveness and quality adaptation functionality over a wide range of network conditions. The advent of 5G networks with Mobile Edge Computing infrastructure presents unforeseen possibilities for shifting bitrate decision-making processes closer to the end user by way of advanced edge-hosted control mechanisms. Multi-armed bandit algorithms used at network edge points exhibit better performance attributes than traditional cloud-based setups, allowing real-time adaptation of quality with lower decision latencies and better user experience metrics. Rigorous experimental assessment using trace-driven emulation on twenty thousand real-world streaming sessions confirms significant performance gains such as startup time savings, rebuffering event reductions, and power efficiency gains through smart quality selection algorithms. Edge-deployed adaptive bitrate controllers take advantage of proximity benefits to attain proactive quality adaptation that avoids buffer starvation incidents while delivering optimal visual quality under changing network conditions. The economic model underpinning edge infrastructure deployment illustrates a strong return on investment through enhanced user retention, adoption of premium service tiers, and lower operational expenses than conventional centralized architectures. Privacy and regulatory benefits are obtained through local data processing that preserves geographic data location and reduces cross-border transfers of information. Edge computing platforms facilitate distributed processing resources that provide stronger protection for privacy through less centralized data collection and better regulatory compliance with multiple jurisdictions.
Downloads
Published
Issue
Section
License

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.