Performance and Energy Trade-Off Analysis of Hierarchical Federated Learning for Plant Disease Classification

TL;DR

This paper analyzes the trade-offs between accuracy and energy efficiency in hierarchical federated learning architectures for plant disease classification, proposing an optimization framework for deployment in IoT environments.

Contribution

It introduces a power- and energy-aware optimization framework for hierarchical federated learning, evaluating various CNN architectures and aggregation strategies for resource-efficient plant disease detection.

Findings

Different model-aggregator combinations show distinct performance-energy trade-offs.

Hierarchical federated architecture reduces communication and computational overhead.

Certain configurations achieve high accuracy with lower energy consumption.

Abstract

Early detection of plant diseases is critical for improving crop productivity, while it also facilitates the foundations of precision agriculture. Recent advances in distributed deep learning have enabled plant disease classification models to be trained across geographically distributed agricultural sensing infrastructures. However, deploying such systems in large-scale Internet of Things (IoT) environments, introduces significant challenges related to computational cost, energy consumption, and system efficiency. In this paper, we present a design-space exploration of hierarchical federated learning architectures for plant disease classification, with a particular focus on the trade-offs between predictive performance and energy efficiency. We further introduce a power- and energy-aware optimization framework that enables the systematic evaluation and selection of model-aggregator configurations under varying deployment constraints. The hierarchical federated architecture organizes distributed clients through intermediate aggregation layers, reducing communication and computational overhead. We evaluate multiple convolutional neural network architectures, including EfficientNet-B0, ResNet-50, and MobileNetV3-Large, in combination with different federated aggregation strategies such as FedAvg, FedProx, and FedAvgM. Experimental results demonstrate that different model-aggregator combinations exhibit distinct performance-energy trade-offs. Consequently, we highlight configurations that achieve competitive diagnostic accuracy and significantly reduce system resource requirements.