Revolutionizing QoE-Driven Network Management with Digital Agents in 6G

TL;DR

The paper addresses QoE-driven network management for 6G by incorporating user behavior and environmental context into QoE modeling. It introduces a two-level digital-agent framework where level-one DAs model per-user QoE and predict resource demands, and level-two DAs abstract distributions to guide adaptive network slicing. A novel QoE metric blends QoS with behavioral dynamics and environmental complexity, formalized as $E = S \times I(B, C)$, linking user experience to underlying service quality. A video-streaming case study demonstrates QoE improvements over benchmark schemes, and the authors discuss practical challenges and potential solutions for efficient data collection, scheduling, and deployment of DAs.

Abstract

In this article, we present a digital agent (DA)-assisted network management framework for future sixth generation (6G) networks considering user quality of experience (QoE). A novel QoE metric is defined by incorporating the impact of user behavioral dynamics and environmental complexity on quality of service (QoS). A two-level DA architecture is proposed to assist the QoE-driven network slicing and orchestration. Three potential solutions are presented from the perspectives of DA data collection, resource scheduling, and DA deployment. A case study demonstrates that the proposed framework can effectively improve user QoE compared with benchmark schemes.

Figures (5)

• Figure 1: DA-assisted network management framework.
• Figure 2: QoE factor formation.
• Figure 3: QoE-based resource demand prediction procedure.
• Figure 4: Tailored network resource orchestration process.
• Figure 5: ELA achievable ratio and real-time QoE comparison.