EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Data and channel-aware device selection for split learning in resource-constrained B5G/6 G networks

Cleyber B. dos Reis, Xavier P. Sebastião, Orlavio C. Averú, Renan R. de Oliveira, Maria do Rosário C. Ribeiro, Waldir Moreira, Gerson Geraldo H. Cavalheiro, Antonio Oliveira-Jr

Annals of Telecommunications · 07.08.2026

DOI: 10.1007/s12243-026-01200-w ↗

Kurz gefasst

Fachartikel in der Zeitschrift Annals of Telecommunications, online erschienen am 07.08.2026. Verfasst von Cleyber B. dos Reis und 7 weiteren Autor:innen. Die Arbeit nennt 39 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Universidade Federal De Goiás.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Split learning (SL) is a promising distributed learning approach for fifth-generation/sixth-generation (B5G/6 G) networks. However, its performance is affected by bandwidth constraints and statistical heterogeneity in the data. Prior works have considered either channel-aware or data-aware device selection in isolation, but few jointly consider both under per-round participation limits. This article extends the “A Network Metric-Aware Split Learning Architecture for B5G/6 G Mobile Networks” (SLArch) framework with explicit bandwidth constraints, partial user equipment (UE) participation, and non-independent and identically distributed (non-IID) data modelled via Dirichlet distributions. We introduce four device selection policies: random, channel-quality-based (ChQual), data-quality-based (DataQual), and a joint channel-and-data-aware policy (ChDataQual) weighted by parameters $$\varphi $$ φ and $$\beta $$ β . Simulations with a total of 150 UEs and 10 participants per round demonstrate key network and SL performance results. DataQual achieves the highest accuracy across most slices, reaching 94.8% on Modified National Institute of Standards and Technology (MNIST) under massive machine-type communications (mMTC), 92.5% on Fashion-MNIST under enhanced mobile broadband (eMBB), and 86% on Federated Extended MNIST (FEMNIST) under eMBB. ChDataQual achieves 93.2% on Fashion-MNIST under eMBB, outperforming DataQual by 0.7%. In terms of reliability, ChQual selection achieves the lowest packet loss of 0.059% under ultra-reliable low-latency communications (URLLC). ChDataQual reaches 91.4% accuracy on MNIST under URLLC while reducing packet loss by 58% relative to DataQual. Furthermore, ChQual and ChDataQual exhibit lower and more stable energy consumption and jitter than Random and DataQual. Throughput remains robust across all strategies, showing that integrated channel-data awareness enables reliable, high-performance SL in resource-constrained B5G/6 G environments.

Förderung

  • Universidade Federal De Goiás

Zitieren

APA

Reis, C. B. D., Sebastião, X. P., Averú, O. C., Oliveira, R. R. D., Ribeiro, M. D. R. C., Moreira, W., Cavalheiro, G. G. H., & Oliveira-Jr, A. (2026). Data and channel-aware device selection for split learning in resource-constrained B5G/6 G networks. Annals of Telecommunications. https://doi.org/10.1007/s12243-026-01200-w

Vancouver

Reis CBD, Sebastião XP, Averú OC, Oliveira RRD, Ribeiro MDRC, Moreira W, et al. Data and channel-aware device selection for split learning in resource-constrained B5G/6 G networks. Annals of Telecommunications. 2026. doi:10.1007/s12243-026-01200-w

BibTeX
@article{Reis2026,
  author = {Reis, Cleyber B. dos and Sebastião, Xavier P. and Averú, Orlavio C. and Oliveira, Renan R. de and Ribeiro, Maria do Rosário C. and Moreira, Waldir and Cavalheiro, Gerson Geraldo H. and Oliveira-Jr, Antonio},
  title = {Data and channel-aware device selection for split learning in resource-constrained B5G/6 G networks},
  journal = {Annals of Telecommunications},
  year = {2026},
  doi = {10.1007/s12243-026-01200-w}
}
Kostenlos starten →