Minimizing active nodes in MEC environments: A distributed learning-driven framework for application placement

Application placement in Multi-Access Edge Computing (MEC) must adhere to service level agreements (SLAs), minimize energy consumption, and optimize metrics based on specific service requirements. In distributed MEC system environments, the placement problem also requires consideration of various ty...

Descripción completa

Detalles Bibliográficos
Autores: Torres Pérez , Claudia, Coronado Calero, Estefanía, Cervelló Pastor , Cristina, Palomares , Javier, Carmona Cejudo, Estela, Shuaib Siddiqui , Muhammad
Tipo de recurso: artículo
Fecha de publicación:2025
País:España
Institución:Universidad de Castilla-La Mancha
Repositorio:RUIdeRA. Repositorio Institucional de la UCLM
OAI Identifier:oai:ruidera.uclm.es:10578/44803
Acceso en línea:https://hdl.handle.net/10578/44803
Access Level:acceso abierto
Palabra clave:Application placement
Distributed deep reinforcement learning
MEC
Scalability
id ES_9eb474a877ecd684affcdd2a1a9f291a
oai_identifier_str oai:ruidera.uclm.es:10578/44803
network_acronym_str ES
network_name_str España
repository_id_str
spelling Minimizing active nodes in MEC environments: A distributed learning-driven framework for application placementTorres Pérez , ClaudiaCoronado Calero, EstefaníaCervelló Pastor , CristinaPalomares , JavierCarmona Cejudo, EstelaShuaib Siddiqui , MuhammadApplication placementDistributed deep reinforcement learningMECScalabilityApplication placement in Multi-Access Edge Computing (MEC) must adhere to service level agreements (SLAs), minimize energy consumption, and optimize metrics based on specific service requirements. In distributed MEC system environments, the placement problem also requires consideration of various types of applications with different entry distribution rates and requirements, and the incorporation of varying numbers of hosts to enable the development of a scalable system. One possible way to achieve these objectives is to minimize the number of active nodes in order to avoid resource fragmentation and unnecessary energy consumption. This paper presents a Distributed Deep Reinforcement Learning-based Capacity-Aware Application Placement (DDRL-CAAP) approach aimed at reducing the number of active nodes in a multi-MEC system scenario that is managed by several orchestrators. Internet of Things (IoT) and Extended Reality (XR) applications are considered in order to evaluate close-to-real-world environments via simulation and on a real testbed. The proposed design is scalable for different numbers of nodes, MEC systems, and vertical applications. The performance results show that DDRL-CAAP achieves an average improvement of 98.3% in inference time compared with the benchmark Integer Linear Programming (ILP) algorithm, and a mean reduction of 4.35% in power consumption compared with a Random Selection (RS) algorithm.Elsevier202520252025info:eu-repo/semantics/articleapplication/pdfapplication/pdfhttps://hdl.handle.net/10578/44803reponame:RUIdeRA. Repositorio Institucional de la UCLMinstname:Universidad de Castilla-La ManchaInglésinfo:eu-repo/semantics/openAccessoai:ruidera.uclm.es:10578/448032026-05-27T07:36:41Z
dc.title.none.fl_str_mv Minimizing active nodes in MEC environments: A distributed learning-driven framework for application placement
title Minimizing active nodes in MEC environments: A distributed learning-driven framework for application placement
spellingShingle Minimizing active nodes in MEC environments: A distributed learning-driven framework for application placement
Torres Pérez , Claudia
Application placement
Distributed deep reinforcement learning
MEC
Scalability
title_short Minimizing active nodes in MEC environments: A distributed learning-driven framework for application placement
title_full Minimizing active nodes in MEC environments: A distributed learning-driven framework for application placement
title_fullStr Minimizing active nodes in MEC environments: A distributed learning-driven framework for application placement
title_full_unstemmed Minimizing active nodes in MEC environments: A distributed learning-driven framework for application placement
title_sort Minimizing active nodes in MEC environments: A distributed learning-driven framework for application placement
dc.creator.none.fl_str_mv Torres Pérez , Claudia
Coronado Calero, Estefanía
Cervelló Pastor , Cristina
Palomares , Javier
Carmona Cejudo, Estela
Shuaib Siddiqui , Muhammad
author Torres Pérez , Claudia
author_facet Torres Pérez , Claudia
Coronado Calero, Estefanía
Cervelló Pastor , Cristina
Palomares , Javier
Carmona Cejudo, Estela
Shuaib Siddiqui , Muhammad
author_role author
author2 Coronado Calero, Estefanía
Cervelló Pastor , Cristina
Palomares , Javier
Carmona Cejudo, Estela
Shuaib Siddiqui , Muhammad
author2_role author
author
author
author
author
dc.subject.none.fl_str_mv Application placement
Distributed deep reinforcement learning
MEC
Scalability
topic Application placement
Distributed deep reinforcement learning
MEC
Scalability
description Application placement in Multi-Access Edge Computing (MEC) must adhere to service level agreements (SLAs), minimize energy consumption, and optimize metrics based on specific service requirements. In distributed MEC system environments, the placement problem also requires consideration of various types of applications with different entry distribution rates and requirements, and the incorporation of varying numbers of hosts to enable the development of a scalable system. One possible way to achieve these objectives is to minimize the number of active nodes in order to avoid resource fragmentation and unnecessary energy consumption. This paper presents a Distributed Deep Reinforcement Learning-based Capacity-Aware Application Placement (DDRL-CAAP) approach aimed at reducing the number of active nodes in a multi-MEC system scenario that is managed by several orchestrators. Internet of Things (IoT) and Extended Reality (XR) applications are considered in order to evaluate close-to-real-world environments via simulation and on a real testbed. The proposed design is scalable for different numbers of nodes, MEC systems, and vertical applications. The performance results show that DDRL-CAAP achieves an average improvement of 98.3% in inference time compared with the benchmark Integer Linear Programming (ILP) algorithm, and a mean reduction of 4.35% in power consumption compared with a Random Selection (RS) algorithm.
publishDate 2025
dc.date.none.fl_str_mv 2025
2025
2025
dc.type.none.fl_str_mv info:eu-repo/semantics/article
format article
dc.identifier.none.fl_str_mv https://hdl.handle.net/10578/44803
url https://hdl.handle.net/10578/44803
dc.language.none.fl_str_mv Inglés
language_invalid_str_mv Inglés
dc.rights.none.fl_str_mv info:eu-repo/semantics/openAccess
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/pdf
application/pdf
dc.publisher.none.fl_str_mv Elsevier
publisher.none.fl_str_mv Elsevier
dc.source.none.fl_str_mv reponame:RUIdeRA. Repositorio Institucional de la UCLM
instname:Universidad de Castilla-La Mancha
instname_str Universidad de Castilla-La Mancha
reponame_str RUIdeRA. Repositorio Institucional de la UCLM
collection RUIdeRA. Repositorio Institucional de la UCLM
repository.name.fl_str_mv
repository.mail.fl_str_mv
_version_ 1869414848613318656
score 15,81155

Similares en LA Referencia