Resource Allocation for Semantic‐Driven Multibeam Satellite Systems via Deep Reinforcement Learning
Internet Technology Letters · 2026-11 · 9 (6)
DOI: 10.1002/itl2.70374 ↗Kurz gefasst
Fachartikel in der Zeitschrift Internet Technology Letters, online erschienen am 09.11.2026. Verfasst von Jianpeng Xu und 2 weiteren Autor:innen. Die Arbeit nennt 18 Literaturangaben. Gefördert von: Natural Science Foundation of Hebei Province.
Abstract (Originalsprache)
This paper focuses on resource allocation in a semantic‐driven multibeam geostationary earth orbit (GEO) satellite system. The sum semantic rate (SSR) is maximized by jointly designing the subchannel and power allocation, while satisfying the semantic similarity. Since the optimization problem is non‐convex and mathematically intractable, an algorithm based on deep reinforcement learning (DRL), specifically the softmax deep double deterministic policy gradients with prioritized experience replay (SD3‐PER) algorithm, is proposed to maximize the reward through sustained interaction with the semantic‐driven multibeam GEO satellite environment. Simulation findings demonstrate that the proposed scheme can achieve higher SSR than the benchmark algorithms.
Förderung
- Natural Science Foundation of Hebei Province
Zitieren
APA
Xu, J., Li, S., & Wang, X. (2026). Resource Allocation for Semantic‐Driven Multibeam Satellite Systems via Deep Reinforcement Learning. Internet Technology Letters, 9(6). https://doi.org/10.1002/itl2.70374
Vancouver
Xu J, Li S, Wang X. Resource Allocation for Semantic‐Driven Multibeam Satellite Systems via Deep Reinforcement Learning. Internet Technology Letters. 2026;9(6). doi:10.1002/itl2.70374
BibTeX
@article{Xu2026,
author = {Xu, Jianpeng and Li, Shuang and Wang, Xuehu},
title = {Resource Allocation for Semantic‐Driven Multibeam Satellite Systems via Deep Reinforcement Learning},
journal = {Internet Technology Letters},
year = {2026},
volume = {9},
number = {6},
doi = {10.1002/itl2.70374}
}
EUVIMED