Mobile Network Control with a World Model

2026-07-20

AI Summary

This work addresses the growing complexity of mobile networks by proposing an intelligent, dynamic energy-saving control strategy grounded in world models. The approach learns from historical data to predict how actions influence future network states and incorporates uncertainty estimation to enable robust decision-making. Notably, it allows for dynamic adjustment of optimization objectives without requiring retraining and autonomously generates energy-efficient configurations in closed-loop control that satisfy quality-of-service constraints. Experimental results demonstrate that the method consistently outperforms conventional approaches and reinforcement learning baselines—both in simulation and on real-world network data—achieving a superior trade-off between energy efficiency and service quality.

Abstract

The increasing complexity of mobile networks necessitates intelligent and dynamic control strategies for efficient, energy-conserving management. We propose a world model-based approach for network control that enables adaptive configuration of crucial parameters. The world model is trained from historical data and predicts the impact of its actions on future network states. Our controller leverages the model's uncertainty estimate to robustly find optimal network configuration changes. Furthermore, the optimization objective can be changed dynamically without model retraining. We demonstrate the effectiveness of the approach in simulated closed-loop control of a mobile network energy-saving feature. Our results show improved performance in balancing energy savings with quality of service, compared to traditional methods and reinforcement learning approaches. Finally, we show the world model performance on real network data from, and evaluate counterfactual actions proposed by the controller under various throughput constraints.

Problem

• mobile network control
• energy efficiency
• quality of service
• dynamic optimization
• world model

Innovation

• world model
• mobile network control
• uncertainty estimation
• dynamic optimization
• energy-saving