Centralized Multi-Objective Client–Gateway Association for Sidelink-Assisted Cellular Offloading
International Journal of Advances in Engineering and Pure Sciences · 30.06.2026 · 38 (2) · 365-385
DOI: 10.7240/jeps.1846009 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Advances in Engineering and Pure Sciences, online erschienen am 30.06.2026. Verfasst von Kübra Uludağ und 1 weiteren Autor:innen. Die Arbeit nennt 37 Literaturangaben.
Abstract (Originalsprache)
The rapid growth of mobile data and user-centric offloading services (e.g., Wi-Fi or sidelink-based relays) requires association mechanisms that go beyond simple signal-strength-based attachment. In this paper, we study a centralized gateway–client association problem in a cellular system with sidelink-assisted offloading, instantiated on a single-cell LTE network with Wi-Fi-based device-to-device connectivity. We model the Multi-Objective Capacitated Client-Gateway Association Problem (MO-CGAP), where a controller assigns each client to at most one gateway, subject to gateway capacity and connectivity constraints, while (i) maximizing the number of connected clients, (ii) maximizing the aggregate achieved data rate, and (iii) minimizing the number of fully loaded gateways to promote load balancing. We show that MO-CGAP is NP-hard even for a single-objective throughput-maximization variant,which motivates heuristic solution methods. We design two metaheuristic algorithms tailored to MO-CGAP: a geneticalgorithm (GA) with a problem-specific encoding, repair, and lexicographic fitness design, and a simulated annealing (SA) scheme that uses the same three-tier objective prioritization. Both are evaluated against a distributed RSSI-based association baseline under realistic cellular downlink and sidelink abstractions, across four spatial scenarios with varying hotspot clustering and different client loads. Numerical results show that GA and SA consistently achieve higher connectivity than the RSSI-based baseline and, for moderate client densities, also improve aggregate throughput and reduce the number of fully loaded gateways by distributing clients more evenly. GA typically yields slightly better throughput and fairness than SA. Although the numerical study is LTE-based, the problem formulation and solution framework are largely radio-access-agnostic and can be adapted to modern OFDMA systems such as 5G NR and NR sidelink-based relay architectures.
Zitieren
APA
Uludağ, K., & Korçak, Ö. (2026). Centralized Multi-Objective Client–Gateway Association for Sidelink-Assisted Cellular Offloading. International Journal of Advances in Engineering and Pure Sciences, 38(2), 365-385. https://doi.org/10.7240/jeps.1846009
Vancouver
Uludağ K, Korçak Ö. Centralized Multi-Objective Client–Gateway Association for Sidelink-Assisted Cellular Offloading. International Journal of Advances in Engineering and Pure Sciences. 2026;38(2):365-385. doi:10.7240/jeps.1846009
BibTeX
@article{Uluda2026,
author = {Uludağ, Kübra and Korçak, Ömer},
title = {Centralized Multi-Objective Client–Gateway Association for Sidelink-Assisted Cellular Offloading},
journal = {International Journal of Advances in Engineering and Pure Sciences},
year = {2026},
volume = {38},
number = {2},
pages = {365--385},
doi = {10.7240/jeps.1846009}
}
EUVIMED