Article (Scientific journals)
Deep Q-Learning-Based Handover for Spectral Coexistence Between Feeder and User Links in LEO Satellite Networks
OLIVEIRA KUHFUSS DE MENDONÇA, Marcele; LAGUNAS, Eva; Grotz, Joel et al.
2025In IEEE Open Journal of the Communications Society, 6, p. 6777 - 6791
Peer Reviewed verified by ORBi
 

Files


Full Text
Deep_Q-Learning-Based_Handover_for_Spectral_Coexistence_Between_Feeder_and_User_Links_in_LEO_Satellite_Networks.pdf
Author postprint (3.94 MB)
Download

All documents in ORBilu are protected by a user license.

Send to



Details



Keywords :
deep Q-network (DQN); handover; interference management; reinforcement learning; Satellite communication; Deep Q-network; Earth orbits; Feeder link; Hand over; Interference management; Ka band; Medium earth orbit satellites; Reinforcement learnings; Satellite communications; Spot beams; Computer Networks and Communications
Abstract :
[en] Deploying feeder links (i.e., between satellites and ground stations) at Ka-band is becoming a popular option among satellite operators thanks to its consolidated technology maturity level. While using Ka-band for feeder links was not a major problem in geostationary satellites, its application in Low-Earth Orbit (LEO) and Medium-Earth Orbit (MEO) satellites brings new challenges due to the fast mobility of LEO/MEO platforms. In particular, the time-varying projection of the satellite spot-beam may result in overlapping with user link spot-beams and create inter-beam interference problems. Traditional beamforming strategies, while theoretically effective in mitigating interference by dynamically adjusting beam shapes, are often impractical due to the required recalculation speed and complexity overhead. In this work, we target a simpler solution based on a smart user-satellite assignment, i.e., handover strategy. A simple handover scheme offers an alternative but is limited by its reliance on accurate, instantaneous Carrier-to-Interference (C/I) values and its inability to adapt to dynamic environmental changes. To address these limitations, this paper proposes a novel handover scheme based on Deep Q-Network (DQN) reinforcement learning. The DQN-based approach enables adaptive learning and optimization of handover decisions based on real-time network conditions, thereby minimizing interference events and enhancing overall network performance. This method improves scalability, resource utilization, and robustness against prediction errors compared to traditional and conventional handover strategies. Comprehensive simulation results are presented, showcasing the proposed scheme’s effectiveness in various scenarios and highlighting its advantages over existing methods. The findings suggest that the DQN-based approach offers a promising solution for managing feeder-user link interference in LEO/MEO satellite communications, ensuring robust and efficient connectivity.
Disciplines :
Electrical & electronics engineering
Author, co-author :
OLIVEIRA KUHFUSS DE MENDONÇA, Marcele  ;  University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > SigCom
LAGUNAS, Eva  ;  University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > SigCom
Grotz, Joel ;  SES S.A., Château de Betzdorf, Betzdorf, Luxembourg
Pérez-Neira, Ana ;  Space and Resilient Communications and Systems, Centre Tecnologic de Telecomunicacions de Catalunya (CTTC/CERCA), Castelldefels, Spain ; Departament de Teoria del Senyal i Comunicacions, Universitat Politecnica de Catalunya, Barcelona, Spain ; ICREA Academia, Barcelona, Spain
CHATZINOTAS, Symeon  ;  University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > SigCom
External co-authors :
yes
Language :
English
Title :
Deep Q-Learning-Based Handover for Spectral Coexistence Between Feeder and User Links in LEO Satellite Networks
Publication date :
20 August 2025
Journal title :
IEEE Open Journal of the Communications Society
eISSN :
2644-125X
Publisher :
Institute of Electrical and Electronics Engineers Inc.
Volume :
6
Pages :
6777 - 6791
Peer reviewed :
Peer Reviewed verified by ORBi
Funders :
Ministerio de Ciencia e Innovación
ICREA Academia program
Luxembourg National Research Fund (FNR
Funding text :
This work was supported in part by the Luxembourg National Research Fund (FNR) through the Project SmartSpace under Grant C21/IS/16193290; in part by the Spanish Ministry of Science and Innovation through the Project IRENE funded by MCIN/AEI/10.13039/501100011033 under Grant PID2020-115323RB-C31; and in part by the ICREA Academia Program funded by the Catalan Government.
Available on ORBilu :
since 17 November 2025

Statistics


Number of views
9 (1 by Unilu)
Number of downloads
4 (0 by Unilu)

Scopus citations®
 
0
Scopus citations®
without self-citations
0
OpenCitations
 
0
OpenAlex citations
 
0
WoS citations
 
0

Bibliography


Similar publications



Contact ORBilu