OrchNAS: Orchestrated Neural Architecture Search Service for Personalised Federated Edge Intelligence

2026-07-24

AI Summary

This work addresses the challenge of jointly optimizing personalization, energy efficiency, and resource constraints in federated learning across heterogeneous edge devices. To this end, the authors propose a server-side neural architecture search framework that integrates energy-aware global architecture representation learning, progressive greedy pruning for personalized subnet selection, and a primal-dual optimization-based adaptive update mechanism to enforce energy consumption constraints. This approach automatically deploys customized models while respecting device-specific limitations on computation, memory, and energy. Experimental results on both real-world and benchmark datasets demonstrate that the proposed method significantly reduces energy consumption, effectively adapts to diverse edge hardware, and maintains competitive model accuracy.

Abstract

We propose OrchNAS, an energy-aware, personalised, federated edge intelligence framework that leverages a Neural Architecture Search Service to automatically design service-adaptive models for heterogeneous edge environments. The framework orchestrates the architecture search process on a server-side NAS service, enabling edge services to derive personalised architectures under device-level energy, computation, and memory constraints. We introduce an energy-aware global architecture search mechanism that learns a compact global representation across heterogeneous services. We develop an energy-efficient architecture selection mechanism that enables each service to derive a personalised subnet that satisfies its resource constraints via a progressive, greedy, energy-aware pruning strategy. We propose an energy-efficient personalised model optimisation scheme that updates service-adaptive parameters while preserving global representations, where a primal-dual optimisation mechanism enforces strict energy budgets during architecture adaptation. Experiments on real-world and benchmark datasets demonstrate the effectiveness of the proposed approach.

Problem

• Neural Architecture Search
• Federated Edge Intelligence
• Energy-aware
• Personalised Models
• Heterogeneous Edge Environments

Innovation

• Neural Architecture Search
• Federated Edge Intelligence
• Energy-aware Optimization
• Personalized Subnet
• Primal-dual Optimization