Optimizing STAR-IRS-Assisted Hybrid Satellite–Terrestrial Relay Networks With CoMP and RSMA: A Deep Reinforcement Learning Approach
IEEE Open Journal of the Computer Society · 2026 · 7 · 1619-1633
DOI: 10.1109/ojcs.2026.3745589 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Open Journal of the Computer Society, 2026. Verfasst von Yaser Azimi. Die Arbeit nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Hybrid satellite–terrestrial relay networks (HSTRNs) have emerged as a promising architecture to deliver broad coverage and high data rates in beyond-5G networks. This paper introduces a downlink framework that combines simultaneously transmitting and reflecting intelligent reflecting surfaces (STAR-IRS), coordinated multi-point (CoMP) transmission, and rate-splitting multiple access (RSMA) to improve spectral and energy efficiency in infrastructure-diverse urban environments. The STAR-IRS reflects and transmits the satellite signal toward a ground base station (BS) and uncrewed aerial vehicle (UAV) relays; the BS and selected UAV relays then employ RSMA, with CoMP used on the relay side. The resulting design couples STAR-IRS phase-shift configuration with RSMA power and rate allocation under predefined STAR-IRS energy-splitting coefficients. We address it using (i) a Lagrangian-dual resource-allocation method with closed-form primal updates and (ii) an asynchronous advantage actor–critic (A3C) method for STAR-IRS phase-shift configuration. To improve scalability, the STAR-IRS is partitioned into sub-surfaces with common phase shifts and predefined amplitude coefficients. Under identical simulated conditions, A3C converges faster and more stably than the implemented deep deterministic policy gradient (DDPG) baseline. The framework is further evaluated for different STAR-IRS sizes, CoMP relay counts, transmit-power budgets, bandwidths, relay heights, and SIC thresholds.
Zitieren
APA
Azimi, Y. (2026). Optimizing STAR-IRS-Assisted Hybrid Satellite–Terrestrial Relay Networks With CoMP and RSMA: A Deep Reinforcement Learning Approach. IEEE Open Journal of the Computer Society, 7, 1619-1633. https://doi.org/10.1109/ojcs.2026.3745589
Vancouver
Azimi Y. Optimizing STAR-IRS-Assisted Hybrid Satellite–Terrestrial Relay Networks With CoMP and RSMA: A Deep Reinforcement Learning Approach. IEEE Open Journal of the Computer Society. 2026;7:1619-1633. doi:10.1109/ojcs.2026.3745589
BibTeX
@article{Azimi2026,
author = {Azimi, Yaser},
title = {Optimizing STAR-IRS-Assisted Hybrid Satellite–Terrestrial Relay Networks With CoMP and RSMA: A Deep Reinforcement Learning Approach},
journal = {IEEE Open Journal of the Computer Society},
year = {2026},
volume = {7},
pages = {1619--1633},
doi = {10.1109/ojcs.2026.3745589}
}
EUVIMED