This paper addresses challenges in UE scheduling for multi-cell wireless networks with federated edge learning, proposing a distributed optimization framework leveraging ADMM and MARL for scalable, low-latency coordination without centralized control.
Key findings
Proposes FedSchedule, a distributed optimization framework combining ADMM and MARL for UE scheduling.
Theoretical convergence guarantees are provided, considering wireless channel conditions and scheduling decisions.
Develops an interference-aware coordination mechanism for collaborative inter-cell interference management.
Simulations show significant improvements in convergence speed, energy efficiency, and fairness compared to baselines.
Limitations & open questions
The paper does not discuss the scalability of the proposed method for very large networks.
The impact of varying data distributions across devices on the proposed framework is not fully explored.