Implementing an Optimal Rate Allocation Tuned to the User Quality of Experience

TL;DR

This paper presents a centralized convex optimization approach for optimal bandwidth allocation in real-world mobile networks, significantly improving user QoE while reducing bandwidth usage and costs.

Contribution

It introduces a novel rate allocation method tailored to user QoE, demonstrated through real-world experiments on a live mobile network.

Findings

Enhanced user QoE with fine-grained temporal resource allocation

Reduced total bandwidth consumption in the network

Lowered operational costs through optimized rate assignment

Abstract

Optimal resource allocation elegantly kaizens bandwidth utilization in present-day communications systems carrying distinctive traffic types with specific quality of service (QoS) requirements, whose fulfillment may elevate users' quality of Experience (QoE). This paper investigates the QoE of users running real-life real-time and delay-tolerant applications by implementing an Internet-connected real-world mobile network which hosts a node with a centralized convex resource allocation optimization algorithm to calculate and enforce an optimal bandwidth distribution. The experiments show that leveraging the rate assignment approach escalates the real-life network traffic QoE through a fine-grained temporal resource allocation pattern which plummets the total bandwidth consumption and the cost of employing the services.