Fast and efficient energy-oriented cell assignment in heterogeneous networks
Ver/ Abrir
Registro completo
Mostrar el registro completo DCAutoría
Rubio-Loyola, Javier; Aguilar-Fuster, Christian; Díez Fernández, Luis Francisco; Agüero Calvo, Ramón; Luis-Gorricho, Juan; Serrat Fernández, JoanFecha
2020-07Derechos
© Springer. This is a post-peer-review, pre-copyedit version of an article published in Wireless Networks. The final authenticated version is available online at: https://doi.org/10.1007/s11276-019-02047-x
Publicado en
Wireless Networks, 2020, 26(5), 3119-3137
Editorial
Springer Netherlands
Enlace a la publicación
Palabras clave
Cell assignment
Resource allocation
Metaheuristic
Energy efficiency
Cellular networks
Heterogeneous networks
Dense networks
Resumen/Abstract
The cell assignment problem is combinatorial, with increased complexity when it is tackled considering resource allocation. This paper models joint cell assignment and resource allocation for cellular heterogeneous networks, and formalizes cell assignment as an optimization problem. Exact algorithms can find optimal solutions to the cell assignment problem, but their execution time increases drastically with realistic network deployments. In turn, heuristics are able to find solutions in reasonable execution times, but they get usually stuck in local optima, thus failing to find optimal solutions. Metaheuristic approaches have been successful in finding solutions closer to the optimum one to combinatorial problems for large instances. In this paper we propose a fast and efficient heuristic that yields very competitive cell assignment solutions compared to those obtained with three of the most widely-used metaheuristics, which are known to find solutions close to the optimum due to the nature of their search space exploration. Our heuristic approach adds energy expenditure reduction in its algorithmic design. Through simulation and formal statistical analysis, the proposed scheme has been proved to produce efficient assignments in terms of the number of served users, resource allocation and energy savings, while being an order of magnitude faster than metaheuritsic-based approaches.