An Explainable Artificial Intelligence Framework for Quality-Aware IoE Service Delivery

XAI improves network transparency: The framework uses Shapley values to explain how different contextual factors (people, data, process, and things) influence IoE service quality, helping network providers understand why the AI makes specific decisions rather than treating it as a "black box."

Significant performance gains demonstrated: The approach achieved uplink improvement rates of 42.43% (AdaBoost) and 16.32% (Extra Trees), with downlink improvements reaching 28.57%, showing that explainability doesn't compromise—and may enhance—system performance.

Addresses a critical gap in autonomous networks: As 6G networks increasingly rely on AI for autonomous control, this framework tackles the problem that AI models often lack interpretability, which can undermine trust and adoption by service providers who need to understand network behavior.

Contextual understanding enables better decisions: By analyzing relationships between contextual metrics (network dynamics, user characteristics, data patterns, and connected devices), the system can reconfigure network parameters more intelligently, moving beyond one-size-fits-all approaches.

Ensemble methods balance accuracy and interpretability: Using multiple regression models together (rather than single complex models) allows the system to maintain both predictive accuracy and the ability to explain which factors matter most for service quality decisions.