Ponències/Comunicacions de congressos
http://hdl.handle.net/2117/2437
2024-03-29T01:28:21Z
2024-03-29T01:28:21Z
Addressing scalability for real-time multiuser holo-portation: introducing and assessing a Multipoint Control Unit (MCU) for volumetric video
Fernández Langa, Sergi
Montagud Climent, Mario
Rincón Rivera, David
Moragues, Jaume
Cernigliaro, Gianluca
http://hdl.handle.net/2117/404805
2024-03-18T11:27:09Z
2024-03-18T11:18:57Z
Addressing scalability for real-time multiuser holo-portation: introducing and assessing a Multipoint Control Unit (MCU) for volumetric video
Fernández Langa, Sergi; Montagud Climent, Mario; Rincón Rivera, David; Moragues, Jaume; Cernigliaro, Gianluca
Scalability, interoperability, and cost efficiency are key remaining challenges to successfully providing real-time holo-portation (and Metaverse-like) services. This paper, for the first time, presents the design and integration of a Multipoint Control Unit (MCU) in a pioneering real-time holo-portation platform, supporting realistic and volumetric user representations (i.e., 3D holograms), with the aim of overcoming such challenges. The feasibility and implications of adopting such an MCU, in comparison with state-of-the-art architectural approaches, are assessed through experimentation in two different deployment setups, by iteratively increasing the number of concurrent users in shared sessions. The obtained results are promising, as it is empirically proved that the newly adopted stream multiplexing together with the novel per-client and per-frame Volumetric Video (VV) processing optimization features provided by the MCU allow increasing the number of concurrent users, while: (i) significantly reducing resources consumption metrics (e.g., CPU, GPU, bandwidth) and frame rate degradation on the client side; and (ii) keeping the end-to-end latency within acceptable limits.
2024-03-18T11:18:57Z
Fernández Langa, Sergi
Montagud Climent, Mario
Rincón Rivera, David
Moragues, Jaume
Cernigliaro, Gianluca
Scalability, interoperability, and cost efficiency are key remaining challenges to successfully providing real-time holo-portation (and Metaverse-like) services. This paper, for the first time, presents the design and integration of a Multipoint Control Unit (MCU) in a pioneering real-time holo-portation platform, supporting realistic and volumetric user representations (i.e., 3D holograms), with the aim of overcoming such challenges. The feasibility and implications of adopting such an MCU, in comparison with state-of-the-art architectural approaches, are assessed through experimentation in two different deployment setups, by iteratively increasing the number of concurrent users in shared sessions. The obtained results are promising, as it is empirically proved that the newly adopted stream multiplexing together with the novel per-client and per-frame Volumetric Video (VV) processing optimization features provided by the MCU allow increasing the number of concurrent users, while: (i) significantly reducing resources consumption metrics (e.g., CPU, GPU, bandwidth) and frame rate degradation on the client side; and (ii) keeping the end-to-end latency within acceptable limits.
Distributed learning for application placement at the edge minimizing active nodes
Torres Pérez, Claudia
Coronado Calero, Estefanía
Cervelló Pastor, Cristina
Camargo Barragán, Juan Sebastián
Siddiqui, Muhammad Shuaib
http://hdl.handle.net/2117/403863
2024-03-06T12:50:17Z
2024-03-06T12:45:50Z
Distributed learning for application placement at the edge minimizing active nodes
Torres Pérez, Claudia; Coronado Calero, Estefanía; Cervelló Pastor, Cristina; Camargo Barragán, Juan Sebastián; Siddiqui, Muhammad Shuaib
The main goal of application placement in Multi-Access Edge Computing (MEC) is to map their requirements to the infrastructure for desired Service Level Agreement (SLA). In highly distributed infrastructures in beyond 5G and 6G networks, meeting this need and minimizing energy use are crucial. Fo-cusing solely on meeting SLA requirements can lead to resource fragmentation and reduced energy efficiency, as nodes utilize only a small portion of their resources. Furthermore, when multiple orchestrators govern MEC nodes, achieving optimal efficiency becomes a more complex challenge. This paper addresses the application placement problem by employing distributed deep reinforcement learning to efficiently minimize the overall cost of active MEC nodes in a distributed scenario involving multiple MEC systems. Our technique reduces the number of active nodes maintaining an average accuracy of up to 98%, meets SLA requirements, and is scalable for hosting several MEC nodes.
© 2023 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes,creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.
2024-03-06T12:45:50Z
Torres Pérez, Claudia
Coronado Calero, Estefanía
Cervelló Pastor, Cristina
Camargo Barragán, Juan Sebastián
Siddiqui, Muhammad Shuaib
The main goal of application placement in Multi-Access Edge Computing (MEC) is to map their requirements to the infrastructure for desired Service Level Agreement (SLA). In highly distributed infrastructures in beyond 5G and 6G networks, meeting this need and minimizing energy use are crucial. Fo-cusing solely on meeting SLA requirements can lead to resource fragmentation and reduced energy efficiency, as nodes utilize only a small portion of their resources. Furthermore, when multiple orchestrators govern MEC nodes, achieving optimal efficiency becomes a more complex challenge. This paper addresses the application placement problem by employing distributed deep reinforcement learning to efficiently minimize the overall cost of active MEC nodes in a distributed scenario involving multiple MEC systems. Our technique reduces the number of active nodes maintaining an average accuracy of up to 98%, meets SLA requirements, and is scalable for hosting several MEC nodes.
Virtual Satellite Network Simulator (VSNeS): a simulation engine to virtualize non-terrestrial networks
Ruiz de Azua, Joan Adria
López Fernandez, Uriel
Ávila Acosta, José Luis
Rincón Rivera, David
http://hdl.handle.net/2117/400869
2024-02-11T05:13:20Z
2024-02-02T11:52:31Z
Virtual Satellite Network Simulator (VSNeS): a simulation engine to virtualize non-terrestrial networks
Ruiz de Azua, Joan Adria; López Fernandez, Uriel; Ávila Acosta, José Luis; Rincón Rivera, David
The space has been populated by a wide range of satellite systems from governmental and private space entities. Monolithic satellites have been ruling the space by providing a custom design that accomplishes a specific mission. However, novel user demands emerged requiring global coverage, low revisit time, and ubiquitous service. The possibility to integrate in-orbit infrastructure to support current mobile communications systems is being discussed persistently in the last years. Specifically, the concept of deploying networks composed of aircraft and spacecraft as nodes, so-called Non-Terrestrial Networks, have emerged as a potential architecture to satisfy this new demand. This novel concept enables to investigate in mobile technologies in space infrastructure. It is the case of the Software-Defined Satellite, which aims at managing in-orbit infrastructure by using Software-Defined Network techniques. However,
simulation engines are required to enhance novel developments in this domain. This work presents the Virtual Satellite Network Simulator, a novel simulation engine capable to represent satellites and ground nodes in virtual machines and deploy a virtual network that represents the channel effects and dynamics. The concept, requirements and design of the software is presented in detail, remarking each component and their functionalities. Additionally, results of the performance evaluation are discussed, demonstrating the viability of the system to deploy large constellations. Finally, a realistic scenario is deployed demonstrating its usage to enhance future developments.
2024-02-02T11:52:31Z
Ruiz de Azua, Joan Adria
López Fernandez, Uriel
Ávila Acosta, José Luis
Rincón Rivera, David
The space has been populated by a wide range of satellite systems from governmental and private space entities. Monolithic satellites have been ruling the space by providing a custom design that accomplishes a specific mission. However, novel user demands emerged requiring global coverage, low revisit time, and ubiquitous service. The possibility to integrate in-orbit infrastructure to support current mobile communications systems is being discussed persistently in the last years. Specifically, the concept of deploying networks composed of aircraft and spacecraft as nodes, so-called Non-Terrestrial Networks, have emerged as a potential architecture to satisfy this new demand. This novel concept enables to investigate in mobile technologies in space infrastructure. It is the case of the Software-Defined Satellite, which aims at managing in-orbit infrastructure by using Software-Defined Network techniques. However,
simulation engines are required to enhance novel developments in this domain. This work presents the Virtual Satellite Network Simulator, a novel simulation engine capable to represent satellites and ground nodes in virtual machines and deploy a virtual network that represents the channel effects and dynamics. The concept, requirements and design of the software is presented in detail, remarking each component and their functionalities. Additionally, results of the performance evaluation are discussed, demonstrating the viability of the system to deploy large constellations. Finally, a realistic scenario is deployed demonstrating its usage to enhance future developments.
Diseño de una herramienta para evaluar la adecuación de una vivienda a sus habitantes y priorizar actuaciones de mejora. Un ejercicio de Diseño Inclusivo
Guasch Murillo, Daniel
León Abarca, Olga
Puig Badía, Laia
Lusilla Roigé, Soraia
Domingo Mora, Montse
http://hdl.handle.net/2117/400267
2024-01-25T12:20:14Z
2024-01-25T12:16:09Z
Diseño de una herramienta para evaluar la adecuación de una vivienda a sus habitantes y priorizar actuaciones de mejora. Un ejercicio de Diseño Inclusivo
Guasch Murillo, Daniel; León Abarca, Olga; Puig Badía, Laia; Lusilla Roigé, Soraia; Domingo Mora, Montse
El artículo describe los trabajos iniciales realizados por la FEPCCAT y la UPC en el marco del proyecto “Rumbo: hacia un modelo de autonomía personal conectada e inclusiva”. El objetivo del artículo es describir la metodología y el diseño de una herramienta digital para valorar la adecuación de una vivienda a sus habitantes y priorizar actuaciones de mejora en esta. Se describe la problemática en la elección de los elementos e indicadores, así como la propuesta de evaluación implementada por la herramienta diseñada.
2024-01-25T12:16:09Z
Guasch Murillo, Daniel
León Abarca, Olga
Puig Badía, Laia
Lusilla Roigé, Soraia
Domingo Mora, Montse
El artículo describe los trabajos iniciales realizados por la FEPCCAT y la UPC en el marco del proyecto “Rumbo: hacia un modelo de autonomía personal conectada e inclusiva”. El objetivo del artículo es describir la metodología y el diseño de una herramienta digital para valorar la adecuación de una vivienda a sus habitantes y priorizar actuaciones de mejora en esta. Se describe la problemática en la elección de los elementos e indicadores, así como la propuesta de evaluación implementada por la herramienta diseñada.
Design and evaluation of a K8s-based system for distributed open-source cellular networks
Palomares Torrecilla, Javier
Coronado, Estefanía
Rincón Rivera, David
Siddiqui, Shuaib
http://hdl.handle.net/2117/392177
2023-07-25T11:30:19Z
2023-07-25T11:24:01Z
Design and evaluation of a K8s-based system for distributed open-source cellular networks
Palomares Torrecilla, Javier; Coronado, Estefanía; Rincón Rivera, David; Siddiqui, Shuaib
Virtualization in cellular networks is one of the key areas of research where technologies, infrastructure and challenges are rapidly changing as 5G system architecture demands a paradigm shift. This paper aims to study the viability and the performance of cloud-native infrastructures for hosting network functions. The selected frameworks implement both the 4G and the 5G stacks and their network functions. This work considers a variety of scenarios for enabling the deployment of a distributed and open-source cellular network: a baremetal setup, an all-docker-based setup and the proposed Kubernetes setup. Moreover, an analysis of the impact that the Radio Access Network (RAN) and the Core Network (CN) have on computational resource utilization is presented as the network conditions vary. The design proposed in this work has been validated and analyzed using the proposed prototype and testbed. This paper proposes a design to increase resource usage flexibility and performance and reduction of deployment time. The analysis of the gathered data reveals that the deployments of containerized cellular networks display better performance in terms of flexibility, low startup times, and ease of deployment while consuming the same resources as the non-containerized.
2023-07-25T11:24:01Z
Palomares Torrecilla, Javier
Coronado, Estefanía
Rincón Rivera, David
Siddiqui, Shuaib
Virtualization in cellular networks is one of the key areas of research where technologies, infrastructure and challenges are rapidly changing as 5G system architecture demands a paradigm shift. This paper aims to study the viability and the performance of cloud-native infrastructures for hosting network functions. The selected frameworks implement both the 4G and the 5G stacks and their network functions. This work considers a variety of scenarios for enabling the deployment of a distributed and open-source cellular network: a baremetal setup, an all-docker-based setup and the proposed Kubernetes setup. Moreover, an analysis of the impact that the Radio Access Network (RAN) and the Core Network (CN) have on computational resource utilization is presented as the network conditions vary. The design proposed in this work has been validated and analyzed using the proposed prototype and testbed. This paper proposes a design to increase resource usage flexibility and performance and reduction of deployment time. The analysis of the gathered data reveals that the deployments of containerized cellular networks display better performance in terms of flexibility, low startup times, and ease of deployment while consuming the same resources as the non-containerized.
Optimizing resource allocation in long-reach PONs for improved performance in 6G networks
Haastrup, Adebanjo
Zehri, Mohammad Habib
Rincón Rivera, David
Piney da Silva, José Ramón
http://hdl.handle.net/2117/392173
2023-10-15T01:47:43Z
2023-07-25T11:13:33Z
Optimizing resource allocation in long-reach PONs for improved performance in 6G networks
Haastrup, Adebanjo; Zehri, Mohammad Habib; Rincón Rivera, David; Piney da Silva, José Ramón
Next-generation cellular networks, such as 6G, will require high-speed and low-latency connectivity to support advanced services and applications. Passive Optical Networks (PONs) have emerged as a promising solution to meet these requirements. Long-reach PONs (LRPONs) have further extended the reach of PONs, making them ideal for remote and rural areas, as well as providing high-speed connectivity to buildings and homes. LRPONs have the potential to play a critical role in the Radio Access Network (RAN) of any mobile communication system, specifically in Centralized Radio Access Network (C-RAN) architectures. In this paper, we present a new dynamic bandwidth allocation (DBA) algorithm specifically enhanced for LRPONs. The proposed algorithm utilizes a scheduling policy that assigns weight vectors to ONUs based on their distance from the OLT, sorting them accordingly without penalizing any ONU due to their distance. We evaluate the performance of the algorithm through simulations under various traffic scenarios. Our results demonstrate that the proposed algorithm outperforms other algorithms in terms of throughput and queue delay, which ultimately reduces the end-to-end delay especially under heavy traffic loads.
2023-07-25T11:13:33Z
Haastrup, Adebanjo
Zehri, Mohammad Habib
Rincón Rivera, David
Piney da Silva, José Ramón
Next-generation cellular networks, such as 6G, will require high-speed and low-latency connectivity to support advanced services and applications. Passive Optical Networks (PONs) have emerged as a promising solution to meet these requirements. Long-reach PONs (LRPONs) have further extended the reach of PONs, making them ideal for remote and rural areas, as well as providing high-speed connectivity to buildings and homes. LRPONs have the potential to play a critical role in the Radio Access Network (RAN) of any mobile communication system, specifically in Centralized Radio Access Network (C-RAN) architectures. In this paper, we present a new dynamic bandwidth allocation (DBA) algorithm specifically enhanced for LRPONs. The proposed algorithm utilizes a scheduling policy that assigns weight vectors to ONUs based on their distance from the OLT, sorting them accordingly without penalizing any ONU due to their distance. We evaluate the performance of the algorithm through simulations under various traffic scenarios. Our results demonstrate that the proposed algorithm outperforms other algorithms in terms of throughput and queue delay, which ultimately reduces the end-to-end delay especially under heavy traffic loads.
The digital platform for the Unite! alliance: the metacampus
Haji Mohammadali, Farnaz
Alcober Segura, Jesús Ángel
http://hdl.handle.net/2117/392172
2023-07-27T08:00:14Z
2023-07-25T11:09:41Z
The digital platform for the Unite! alliance: the metacampus
Haji Mohammadali, Farnaz; Alcober Segura, Jesús Ángel
Nine European technology universities constitute the Unite! Alliance, a network that creates a new model for a virtual and physical European interuniversity campus through close cooperation, physical and virtual mobility of members, joint programmes, communities that support teaching innovation, and open and entrepreneurial innovation networks. The Metacampus virtual campus platform is an important aspect of this framework. All participants in the Unite! Community can join at the Metacampus, which coexists alongside other universities' virtual campuses. It must become a valuable and useful tool for a wide range of users who are used to very different tools and processes, which is only one of the many difficulties it faces. The Moodle platform was chosen as the base for the work, and it was decided to maximise its features while minimising the amount of new innovations required to achieve the goals. After four years, the number of users who have registered and taken part in the activities that have been suggested enables us to draw the conclusion that the path is the right one for handling a challenge like the design and execution of a multi-university virtual campus.
2023-07-25T11:09:41Z
Haji Mohammadali, Farnaz
Alcober Segura, Jesús Ángel
Nine European technology universities constitute the Unite! Alliance, a network that creates a new model for a virtual and physical European interuniversity campus through close cooperation, physical and virtual mobility of members, joint programmes, communities that support teaching innovation, and open and entrepreneurial innovation networks. The Metacampus virtual campus platform is an important aspect of this framework. All participants in the Unite! Community can join at the Metacampus, which coexists alongside other universities' virtual campuses. It must become a valuable and useful tool for a wide range of users who are used to very different tools and processes, which is only one of the many difficulties it faces. The Moodle platform was chosen as the base for the work, and it was decided to maximise its features while minimising the amount of new innovations required to achieve the goals. After four years, the number of users who have registered and taken part in the activities that have been suggested enables us to draw the conclusion that the path is the right one for handling a challenge like the design and execution of a multi-university virtual campus.
Enabling intelligence inclusiveness in edge to cloud continuum challenges and opportunities
Palomares Torrecilla, Javier
Coronado, Estefanía
Cervelló Pastor, Cristina
Siddiqui, Muhammad Shuaib
http://hdl.handle.net/2117/391973
2023-07-24T07:52:55Z
2023-07-24T07:36:36Z
Enabling intelligence inclusiveness in edge to cloud continuum challenges and opportunities
Palomares Torrecilla, Javier; Coronado, Estefanía; Cervelló Pastor, Cristina; Siddiqui, Muhammad Shuaib
Edge to Cloud Continuum is a concept that integrates cloud computing and cellular networks that has been gaining popularity due to its potential to provide a seamless user experience and address the challenges of managing complex multi-domain networks involving massive IoT devices. Enabling intelligence in the Edge to Cloud Continuum can further enhance its capabilities, offering benefits such as reduced latency, improved scalability, enhanced resource utilization, and increased context awareness. This paper provides insights into the opportunities and challenges of enabling intelligence in Edge to Cloud Continuum, highlighting the potential of this technology. This study presents a comprehensive review of the existing literature on enabling intelligence in Edge to Cloud Continuum, to reach the research questions that will construct the PhD. Various tools and technologies that can be used to integrate intelligence into the Edge to Cloud Continuum system were explored and analyzed. In addition, this study provides a detailed work plan for the upcoming months of the project.
© 2023 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes,creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.
2023-07-24T07:36:36Z
Palomares Torrecilla, Javier
Coronado, Estefanía
Cervelló Pastor, Cristina
Siddiqui, Muhammad Shuaib
Edge to Cloud Continuum is a concept that integrates cloud computing and cellular networks that has been gaining popularity due to its potential to provide a seamless user experience and address the challenges of managing complex multi-domain networks involving massive IoT devices. Enabling intelligence in the Edge to Cloud Continuum can further enhance its capabilities, offering benefits such as reduced latency, improved scalability, enhanced resource utilization, and increased context awareness. This paper provides insights into the opportunities and challenges of enabling intelligence in Edge to Cloud Continuum, highlighting the potential of this technology. This study presents a comprehensive review of the existing literature on enabling intelligence in Edge to Cloud Continuum, to reach the research questions that will construct the PhD. Various tools and technologies that can be used to integrate intelligence into the Edge to Cloud Continuum system were explored and analyzed. In addition, this study provides a detailed work plan for the upcoming months of the project.
La Conceptualització del Metacampus d'Unite!
Alcober Segura, Jesús Ángel
Llorens García, Ariadna
Adam, Araceli
Oller Aubia, Mercè
Prat Farran, Joana d'Arc
Alier Forment, Marc
http://hdl.handle.net/2117/387004
2023-05-04T10:40:18Z
2023-05-04T10:30:22Z
La Conceptualització del Metacampus d'Unite!
Alcober Segura, Jesús Ángel; Llorens García, Ariadna; Adam, Araceli; Oller Aubia, Mercè; Prat Farran, Joana d'Arc; Alier Forment, Marc
L’Aliança Unite!, formada per set universitats tecnològiques europees, és una xarxa que estableix un nou model per a un campus interuniversitari europeu virtual i físic, de col·laboració estreta, mitjançant mobilitat física i virtual dels seus membres, programes conjunts, comunitats promotores d’innovació docent i xarxes d’innovació oberta i emprenedoria. Una peça clau d’aquest entramat és la plataforma de campus virtual Metacampus. Convivint amb els campus virtuals de les altres universitats, el Metacampus té com a funció ser un punt de trobada comuna de tot membre de l’ecosistema Unite! amb reptes importants, com ara esdevenir una eina útil i usable per un conjunt heterogeni d’usuaris, acostumats a formes i processos força diferents. S’ha optat per utilitzar com a base la plataforma Moodle i maximitzar l’ús de les seves funcionalitats, mentre que es redueix al màxim el nombre de nous desenvolupaments que ajudin a aconseguir els objectius establerts. Després de tres mesos en la fase beta, el número d’usuaris registrats i participants en les activitats proposades fins ara permeten concloure que el camí iniciat ara fa mig any és l’adequat per abordar un repte com ara el disseny i la implantació d’un campus virtual multi universitari.
Aquesta publicació s’ha realitzat en el marc del projecte: / Esta publicación se ha realizado en el marco del proyecto: / This publication has been made with the framework of the project: ACEDIM: Avaluació i certificació de la competència digital docent en la formació inicial de mestres: una proposta de model per al sistema universitari català. (ref. 2017ARMIF00031).
2023-05-04T10:30:22Z
Alcober Segura, Jesús Ángel
Llorens García, Ariadna
Adam, Araceli
Oller Aubia, Mercè
Prat Farran, Joana d'Arc
Alier Forment, Marc
L’Aliança Unite!, formada per set universitats tecnològiques europees, és una xarxa que estableix un nou model per a un campus interuniversitari europeu virtual i físic, de col·laboració estreta, mitjançant mobilitat física i virtual dels seus membres, programes conjunts, comunitats promotores d’innovació docent i xarxes d’innovació oberta i emprenedoria. Una peça clau d’aquest entramat és la plataforma de campus virtual Metacampus. Convivint amb els campus virtuals de les altres universitats, el Metacampus té com a funció ser un punt de trobada comuna de tot membre de l’ecosistema Unite! amb reptes importants, com ara esdevenir una eina útil i usable per un conjunt heterogeni d’usuaris, acostumats a formes i processos força diferents. S’ha optat per utilitzar com a base la plataforma Moodle i maximitzar l’ús de les seves funcionalitats, mentre que es redueix al màxim el nombre de nous desenvolupaments que ajudin a aconseguir els objectius establerts. Després de tres mesos en la fase beta, el número d’usuaris registrats i participants en les activitats proposades fins ara permeten concloure que el camí iniciat ara fa mig any és l’adequat per abordar un repte com ara el disseny i la implantació d’un campus virtual multi universitari.
Design of AI-based resource forecasting methods for network slicing
Camargo, Juan Sebastián
Coronado, Estefanía
Gómez, Blas
Rincón Rivera, David
Siddiqui, Shuaib
http://hdl.handle.net/2117/386997
2023-05-04T09:20:26Z
2023-05-04T09:15:57Z
Design of AI-based resource forecasting methods for network slicing
Camargo, Juan Sebastián; Coronado, Estefanía; Gómez, Blas; Rincón Rivera, David; Siddiqui, Shuaib
With the forthcoming of 5G networks, the underlying infrastructure needs to support a higher number of heterogeneous services with different QoS needs than ever. For that reason, 5G inherently provides a way to allocate these services over the same infrastructure through the concept of Network Slicing. However, to maximize revenue and reduce operational costs, a method to proactively adapt the resources assigned to each slice becomes imperative. For that reason, this work presents two
Machine Learning (ML) models, leveraging Long-Short Term Memory (LSTM) and Random Forest algorithms, to forecast the throughput of each slice and adapt accordingly the amount of resources needed. The models are evaluated using NS-3, which has
been integrated with the ML models through a shared memory framework. This enables a closed loop in which the predictions
of the models can be used at run time to introduce changes in the network. Consequently, it makes it able to cope with
the forecasted requirements, eliminating the need for off-line training and resembling better a real-life scenario. The evaluation performed shows the ability of the models to predict the slices’ throughput under various settings and proves that Random Forest
provides up to 26% better results than LSTM.
© 2022 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.
2023-05-04T09:15:57Z
Camargo, Juan Sebastián
Coronado, Estefanía
Gómez, Blas
Rincón Rivera, David
Siddiqui, Shuaib
With the forthcoming of 5G networks, the underlying infrastructure needs to support a higher number of heterogeneous services with different QoS needs than ever. For that reason, 5G inherently provides a way to allocate these services over the same infrastructure through the concept of Network Slicing. However, to maximize revenue and reduce operational costs, a method to proactively adapt the resources assigned to each slice becomes imperative. For that reason, this work presents two
Machine Learning (ML) models, leveraging Long-Short Term Memory (LSTM) and Random Forest algorithms, to forecast the throughput of each slice and adapt accordingly the amount of resources needed. The models are evaluated using NS-3, which has
been integrated with the ML models through a shared memory framework. This enables a closed loop in which the predictions
of the models can be used at run time to introduce changes in the network. Consequently, it makes it able to cope with
the forecasted requirements, eliminating the need for off-line training and resembling better a real-life scenario. The evaluation performed shows the ability of the models to predict the slices’ throughput under various settings and proves that Random Forest
provides up to 26% better results than LSTM.