Perceptron-Based Multi-Criteria Scheduler for 5G and Beyond Networks
IEEE Communications Letters · 2026 · 30 · 2332-2336
DOI: 10.1109/lcomm.2026.3695765 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Communications Letters, 2026. Verfasst von Rajesh Kumar Mahato und 1 weiteren Autor:innen. Die Arbeit nennt 25 Literaturangaben. Gefördert von: ANRF, India.
Abstract (Originalsprache)
Next-generation wireless networks demand intelligent and efficient scheduling mechanisms that can jointly optimise throughput, latency, and fairness while remaining computationally tractable. This letter proposes a Perceptron-Based Multi-Criteria (PB-MC) scheduler, leveraging a lightweight single-layer neural network with online gradient adaptation. The scheduler dynamically categorises users into prime and non-prime classes based on their quality-of-service (QoS) profiles, and fuses normalised key performance indicators—SINR, CQI, buffer occupancy, and throughput—at every transmit time interval (TTI). An $\epsilon $ -constrained fairness mechanism ensures minimum resource allocation for non-prime users, addressing a major limitation of conventional schedulers. 3GPP-compliant simulations demonstrate that the PB-MC scheduler achieves near-optimal fairness (Jain’s Index = 0.96), 21.4% higher aggregate throughput, and 99.2% reliability within 1 ms latency, outperforming the widely used Proportional Fair (PF) scheduler. The proposed approach offers a scalable and practical solution for real-time, fair, and efficient scheduling in 5G and beyond networks.
Förderung
- ANRF, India
Zitieren
APA
Mahato, R. K., & Singh, P. (2026). Perceptron-Based Multi-Criteria Scheduler for 5G and Beyond Networks. IEEE Communications Letters, 30, 2332-2336. https://doi.org/10.1109/lcomm.2026.3695765
Vancouver
Mahato RK, Singh P. Perceptron-Based Multi-Criteria Scheduler for 5G and Beyond Networks. IEEE Communications Letters. 2026;30:2332-2336. doi:10.1109/lcomm.2026.3695765
BibTeX
@article{Mahato2026,
author = {Mahato, Rajesh Kumar and Singh, Prem},
title = {Perceptron-Based Multi-Criteria Scheduler for 5G and Beyond Networks},
journal = {IEEE Communications Letters},
year = {2026},
volume = {30},
pages = {2332--2336},
doi = {10.1109/lcomm.2026.3695765}
}
EUVIMED