Abstract:This paper studies the wireless scheduling design to coordinate the transmissions of (local) model parameters of federated learning (FL) for a swarm of unmanned aerial vehicles (UAVs). The overall goal of the proposed design is to realize the FL training and aggregation processes with a central aggregator exploiting the sensory data collected by the UAVs but it considers the multi-hop wireless network formed by the UAVs. Such transmissions of model parameters over the UAV-based wireless network potentially cause large transmission delays and overhead. Our proposed framework smartly aggregates local model parameters trained by the UAVs while efficiently transmitting the underlying parameters to the central aggregator in each FL global round. We theoretically show that the proposed scheme achieves minimal delay and communication overhead. Extensive numerical experiments demonstrate the superiority of the proposed scheme compared to other baselines.

What problem does this paper attempt to address?

This paper aims to solve the problems of transmission delay and communication overhead faced by unmanned aerial vehicle (UAV) swarms when performing federated learning (FL) tasks. Specifically, the paper focuses on how to effectively schedule wireless communications between UAVs to coordinate the transmission of model parameters without infrastructure support, thereby achieving an efficient and low - latency federated learning training and aggregation process. ### Core Problems of the Paper - **Delay and Overhead Problems**: When implementing federated learning in a UAV swarm, due to the existence of multi - hop communications, the wireless transmission of model parameters may lead to large transmission delays and communication overheads. The goal of the paper is to design a transmission scheduling scheme to minimize these delays and overheads. - **Collaborative Learning in an Infrastructure - free Environment**: In the absence of infrastructure such as ground stations, UAVs need to self - organize to complete data collection, model training, and model aggregation without the support of an external central station. ### Solutions - **Transmission Scheduling Scheme**: The paper proposes a transmission scheduling scheme that realizes the efficient transmission of model parameters by selecting appropriate aggregator nodes and optimizing the transmission paths between UAVs. This scheme can minimize transmission delays and communication overheads. - **Graph - theory Method**: Model the UAV swarm as a graph and select the optimal aggregator node through graph - theory methods (such as breadth - first search), and design a transmission matrix to manage the transmission activities between UAVs. - **Local Aggregation**: Each UAV can first perform local aggregation of model parameters locally and then transmit the aggregated results to the aggregator node, thereby reducing the total number of transmissions and overheads. ### Experimental Verification - **Convergence Performance**: The experimental results show that the proposed scheme has a faster convergence speed than existing baseline methods on two standard datasets, MNIST and CIFAR - 10. - **Delay and Overhead**: By comparing the root node selections under different strategies, the experiments show that the proposed scheme can significantly reduce transmission delays and communication overheads in UAV swarms of different scales. ### Main Contributions - **Theoretical Analysis**: The paper theoretically proves that the proposed scheme can achieve the minimum transmission delay and communication overhead. - **Practical Application**: Verify the effectiveness of the scheme through numerical experiments and demonstrate its superior performance in practical scenarios. ### Conclusion The paper proposes a federated learning transmission scheduling algorithm suitable for UAV swarms, which can achieve efficient model parameter transmission in an infrastructure - free environment and significantly reduce transmission delays and communication overheads. Future research directions may include exploring the impact of topological changes on the federated learning framework.

Delay and Overhead Efficient Transmission Scheduling for Federated Learning in UAV Swarms

Non-orthogonal Multiple Access Assisted Federated Learning for UAV Swarms: an Approach of Latency Minimization

Wireless Federated Learning over UAV-enabled Integrated Sensing and Communication

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

A Federated Learning Latency Minimization Method for UAV Swarms Aided by Communication Compression and Energy Allocation

Energy-Efficient Dynamic Device Scheduling for Over-the-Air Federated Learning in UAV Swarms

Joint Trajectory and Communication Optimization for UAV-Assisted Over-The-Air Federated Learning.

When UAV Meets Federated Learning: Latency Minimization via Joint Trajectory Design and Resource Allocation

UAV Communications for Sustainable Federated Learning

Energy-Efficient Federated Learning Through UAV Edge under Location Uncertainties

UAV Swarm-Assisted Two-Tier Hierarchical Federated Learning

UAV-empowered Vehicular Networking Scheme for Federated Learning in Delay Tolerant Environments

Delay Minimization for Federated Learning Over Wireless Communication Networks

Multi-UAV Hierarchical Intelligent Traffic Offloading Network Optimization Based on Deep Federated Learning

A Fairness-Enhanced Federated Learning Scheduling Mechanism for UAV-Assisted Emergency Communication

Optimizing Data-Driven Federated Learning in UAV Networks

Enhancing the Efficiency of UAV Swarms Communication in 5G Networks Through a Hybrid Split and Federated Learning Approach.

Federated Learning via Unmanned Aerial Vehicle

Asynchronous Federated Learning over Wireless Communication Networks.

Decentralized federated learning methods for reducing communication cost and energy consumption in UAV networks

Experimental Results of Federated Learning in Unmanned Aerial Vehicle Swarm