Federated Learning in the Sky: Joint Power Allocation and Scheduling with UAV Swarms

Abstract

Unmanned aerial vehicle (UAV) swarms must exploit machine learning (ML) in order to execute various tasks ranging from coordinated trajectory planning to cooperative target recognition. However, due to the lack of continuous connections between the UAV swarm and ground base stations (BSs), using centralized ML will be challenging, particularly when dealing with a large volume of data. In this paper, a novel framework is proposed to implement distributed federated learning (FL) algorithms within a UAV swarm that consists of a leading UAV and several following UAVs. Each following UAV trains a local FL model based on its collected data and then sends this trained local model to the leading UAV who will aggregate the received models, generate a global FL model, and transmit it to followers over the intra-swarm network. To identify how wireless factors, like fading, transmission delay, and UAV antenna angle deviations resulting from wind and mechanical vibrations, impact the performance of FL, a rigorous convergence analysis for FL is performed. Then, a joint power allocation and scheduling design is proposed to optimize the convergence rate of FL while taking into account the energy consumption during convergence and the delay requirement imposed by the swarm's control system. Simulation results validate the effectiveness of the FL convergence analysis and show that the joint design strategy can reduce the number of communication rounds needed for convergence by as much as 35% compared with the baseline design.

Federated Learning for Internet of Things: A Comprehensive Survey

“…The simulations confirm that federated learning algorithms can reduce transmission time, motion energy of UAV communications, as well as achieve minimum collision risks in windy environments. Instead of using a central server like in [141], the authors in [142] use a leading UAV as an FL aggregator to manage a swarm of following UAVs in an intra-swarm network. Based on a defined minimum number of communication rounds, the key objective is to jointly optimize both power allocation and scheduling for UAVs, aiming to reduce the FL convergence round, with respect to learning, communication delay, and flying coverage constraints.…”

Federated Machine Learning: Survey, Multi-Level Classification, Desirable Criteria and Future Directions in Communication and Networking Systems

“…In [52], the authors capitalize on federated learning to design aerial-ground air quality sensing framework for fine-grained 3D air quality monitoring and prediction. In [53], the authors address the problem of intermittent connections among UAV swarms and ground base stations, which hinders the adoption of centralized machine learning for executing various tasks (e.g., target recognition, etc.). To do so, a federated learning approach is proposed where each UAV trains a local federated learning model based on its collected data and then forwards the trained model to a leading UAV node, which is responsible for model aggregation.…”

Artificial Intelligence for UAV-Enabled Wireless Networks: A Survey

“…C. FL Solutions for UAVs 1) FL for resource allocation and scheduling In [168], an optimization problem is formulated to design joint power allocation and scheduling for a UAV swarm network. The considered network is composed of one leader UAV and a group of following UAVs.…”