Designing Sustainable Federated Learning as a Service using Neural Architecture Search

Keya Patel; Sajib Mistry; Sheik Fattah; Deepak Kanneganti; Aneesh Krishna; Mufti Mahmud; Monowar Bhuyan

cs.AIAug 14, 2026

Designing Sustainable Federated Learning as a Service using Neural Architecture Search

Authors: Keya Patel, Sajib Mistry, Sheik Fattah, Deepak Kanneganti, Aneesh Krishna, Mufti Mahmud, Monowar Bhuyan

Organizations: School of Electrical Engineering, Computing and Mathematical Sciences Curtin University, Perth, Australia · King Fahd University of Petroleum and Minerals, Saudi Arabia · Umeå University, Umeå, SE-90187, Sweden

Abstract

The sustainability constraints of FLaaS consumers pose significant challenges to maintaining carbon-feasible federated training in FLaaS environments. These constraints often lead to infeasible consumer participation and unstable federated training under hard carbon constraints. We propose a Sustainable Federated Learning as a Service (SFLaaS), a carbon- constrained Neural Architecture Search (NAS) framework for heteroge- neous sustainable constraints. We introduce a requirement-driven search space that transforms consumer sustainability profiles into a feasible architecture region before federated execution. We develop a consumer-level carbon feasibility estimation mechanism to evaluate candidate architectures under dynamic carbon conditions. We propose a sustainable con- sumer scheduling strategy that adaptively selects feasible consumers and allocates local workloads to preserve consumer participation and statistical data coverage. An evolutionary search strategy jointly optimised for predictive performance, consumer feasibility, and participation coverage under hard carbon constraints. Experiments on real-world datasets and a simulated environment demonstrate the effectiveness of the proposed approach.

Figures & tables

Consumer	$\bm{\eta_{i}}$ (GFLOPs/J)	$\bm{CI_{i}}$ (gCO 2 /kWh)	$\bm{B_{i}}$ (gCO 2 )	$\bm{\|D_{i}\|}$ (images)
$c_{1}$ Sydney	2.1	420	800	12,000
$c_{2}$ Singapore	3.4	310	1500	8,000
$c_{3}$ Berlin	4.8	180	2000	15,000
$c_{4}$ Toronto	2.6	250	1200	10,000

Table 1: Consumer constraints in the motivating example.

Figure 2: An Overview of SFLaaS-NAS Framework

Notation	Description
$\alpha\in\mathcal{A}$	Candidate architecture
$R^{*}$	Unified sustainability profile
$\mathcal{A}_{\mathrm{feasible}}$	Feasible architecture set
$b_{i}(t)$	Remaining carbon budget of consumer $i$
$\mathrm{Carbon}_{i}^{t}(\alpha,E_{i}^{t})$	Consumer-side carbon cost
$\Gamma_{i}(\alpha)$	Carbon-feasible participation region

Table 2: Notation and Description

Dataset	Method	Accuracy (%)	Participation (%)	Violation (%)
	Accuracy-Driven NAS	85.8 $\pm$ 0.8	58.3 $\pm$ 2.6	26.7 $\pm$ 2.1
	Efficiency-Aware NAS	83.9 $\pm$ 0.9	74.6 $\pm$ 2.0	12.4 $\pm$ 1.7
	Federated NAS	85.1 $\pm$ 0.7	69.2 $\pm$ 2.4	15.8 $\pm$ 1.5
CIFAR-10	Global Carbon-Aware NAS	84.3 $\pm$ 0.6	78.5 $\pm$ 1.8	9.6 $\pm$ 1.2
	SFLaaS-NAS	86.4 $\pm$ 0.5	94.8 $\pm$ 1.6	1.2 $\pm$ 0.7
	Accuracy-Driven NAS	92.9 $\pm$ 0.6	61.2 $\pm$ 2.4	24.8 $\pm$ 1.8

Table 3: Performance Comparison under Carbon Constraints across Datasets

Method	Participation (%)	Violation (%)
Without Feasibility Filtering	82.1 $\pm$ 2.3	11.4 $\pm$ 1.4
SFLaaS-NAS	94.8 $\pm$ 1.6	1.2 $\pm$ 0.7

Table 4: Impact of Feasibility-Aware Search Space Reduction

Figure 3: (a) Consumer Feasibility Analysis, (b) Impact of Dynamic Scheduling, (c) Evolutionary Search Convergence, and (d) Scalability Analysis

Figure 4: (e) Robustness under Correlated Carbon-Intensity Spikes, and (f) Sensitivity Analysis under Fixed and Tuned Weight Configurations.

Explore similar work

Auto-FL-Research: Agentic Search for Federated Learning Algorithms
Jul 1, 2026Holger R. Roth, Ziyue Xu, Chester Chen +3Automated Algorithm Discovery Autonomous Research Agents
NVIDIA, Santa Clara, USA
Federated Learning by Utility-Constrained Stochastic Aggregation for Improving Rational Participation
May 18, 2026M Yashwanth, Arunabh Singh, Ashok Nayak +2Federated Learning Aggregation Federated Learning
Indian Institute of Science, Bangalore · Indian Institute of Technology Bombay · IIIT Hyderabad
Optimized Federated Knowledge Distillation with Distributed Neural Architecture Search
May 20, 2026Chaimaa Medjadji, Sylvain Kubler, Yves Le Traon +3Communication-Efficient Distributed Training Federated Knowledge Distillation
Interdisciplinary Centre for Security, Reliability and Trust (SnT), University of Luxembourg, Luxembourg · Blekinge Institute of Technology, Sweden · ADSLabs, Umea University, Sweden

Jul 1, 2026, cs.AIJ/K move · Enter open · S save

Auto-FL-Research: Agentic Search for Federated Learning Algorithms

Holger R. Roth, Ziyue Xu, Chester Chen, Daguang Xu+2

NVIDIA, Santa Clara, USA

Automated Algorithm Discovery Autonomous Research Agents Federated Learning

Open paper AskPDF arXiv

Federated learning (FL) research often depends on many small but consequential algorithmic choices: optimizer variants, server aggregation rules, local training schedules, normalization, regularization, and model architecture. These choices are expensive to explore manually and difficult to compare fairly when candidate changes can also alter the FL training or evaluation path. In this work, we present Auto-FL-Research (AFR), a constrained coding-agent workflow for FL algorithmic recipe search. Agents may propose and implement candidate training algorithms, including server aggregation rules, client update schedules, local objectives, and registered model variants, while task profiles fix the mutation surface, compute budget, communication contract, and final model evaluation. Each campaign records candidate scores, runtime, edited files, artifacts, and failure status. We evaluate AFR on five healthcare cross-silo FLamby tasks and on grouped-client profiles for the five fixed LEAF datasets plus the LEAF synthetic task. Five-seed repeat evaluations support gains on four FLamby tasks and five of six LEAF profiles, while also exposing seed-sensitive and search-selected failure cases. Same-budget controls show that several gains correspond to FL-recipe changes, whereas other improvements are recovered by fixed-surface scalar controls or fail under repeat or held-out evaluation. These mixed outcomes are part of the contribution: they show how agent-generated candidates can be separated into repeated FL mechanisms, fixed-surface tuning effects, and selected single-run artifacts.

Abstract

Explore similar work

Auto-FL-Research: Agentic Search for Federated Learning Algorithms

Federated Learning by Utility-Constrained Stochastic Aggregation for Improving Rational Participation

Optimized Federated Knowledge Distillation with Distributed Neural Architecture Search