Sim-to-Real Transfer in Multi-agent Reinforcement Networking for Federated Edge Computing

TL;DR

This paper introduces FedEdge, a high-fidelity simulator for multi-hop federated learning over wireless edge networks, enabling effective sim-to-real transfer and rapid prototyping.

Contribution

The paper presents FedEdge, a novel simulator with realistic physical layer emulation that bridges the gap between simulation and real-world multi-hop federated learning systems.

Findings

FedEdge achieves high fidelity in simulating wireless multi-hop networks.

The simulator enables effective transfer of reinforcement learning policies to real systems.

FedEdge outperforms existing simulation tools in mimicking real-world network dynamics.

Abstract

Federated Learning (FL) over wireless multi-hop edge computing networks, i.e., multi-hop FL, is a cost-effective distributed on-device deep learning paradigm. This paper presents FedEdge simulator, a high-fidelity Linux-based simulator, which enables fast prototyping, sim-to-real code, and knowledge transfer for multi-hop FL systems. FedEdge simulator is built on top of the hardware-oriented FedEdge experimental framework with a new extension of the realistic physical layer emulator. This emulator exploits trace-based channel modeling and dynamic link scheduling to minimize the reality gap between the simulator and the physical testbed. Our initial experiments demonstrate the high fidelity of the FedEdge simulator and its superior performance on sim-to-real knowledge transfer in reinforcement learning-optimized multi-hop FL.