Pasar al contenido principal

Resultados relacionados

Resultados totales (Incluyendo duplicados): 36
Encontrada(s) 1 página(s)

e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

GANSO: Automate Network Slicing at the Transport Network Interconnecting the Edge

  • Takeru Infiesta, José
  • Guimarães, Carlos
  • Contreras Murillo, Luis Miguel
  • Oliva Delgado, Antonio de la
Proceeding of: 2020 IEEE Conference on Network Function Virtualization and Software Defined Networks (NFV-SDN), proceedings, 9 -12 November 2020 // Virtual Conference, Madrid/Leganés, Spain, 5G and Edge computing are two technologies set to impose a paradigm shift from today's traditional networking solutions. In particular, transport networks, which connect distinct computing infrastructures, must guarantee a wide range of performance requirements from coexisting network services. 5G network slicing enables such capability by providing the flexibility to support multiple and isolated virtual networks over the same and shared infrastructure. This paper introduces the GST And Network Slice Operator (GANSO) framework for automating the creation of network slices over SDN architectures, focusing on transport networks interconnecting Edge data centers. To characterise the type of network slice to be deployed, it uses Generic network Slice Templates (GSTs). Initially, five GST attributes are implemented in a proof-of-concept prototype, namely through configurable User Data Access and Rate Limit parameters. It is then validated in a scenario considering the instantiation of network slices over the transport network for different virtual applications hosted across the edge-to-cloud continuum., This work has been (partially) funded by H2020 EU/TW 5G-DIVE (Grant 859881) and H2020 5Growth (Grant 856709). It has been also funded by the Spanish State Research Agency (TRUE5G project, PID2019-108713RB-C52PID2019-108713RB-C52 / AEI / 10.13039/501100011033).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

Demo: vrAIn proof-of-concept: a deep learning approach for virtualized RAN resource control

  • Ayala Romero, José A.
  • García Saavedra, Andrés
  • Gramaglia, Marco
  • Costa-Pérez, Xavier
  • Banchs Roca, Albert
  • Alcaraz, Juan J.
Proceeding of: 25th Annual International Conference on Mobile Computing and Networking (MobiCom'19), October 21-25, 2019, Los Cabos, Mexico., While the application of the NFV paradigm into the network is proceeding full steam ahead, there is still one last milestone to be achieved in this context: the virtualization of the radio access network (vRAN). Due to the very complex dependency between the radio conditions and the computing resources needed to provide the baseband processing functionality, attaining an efficient resource control is particularly challenging. In this demonstration, we will showcase vrAIn, a vRAN dynamic resource controller that employs deep reinforcement learning to perform resource assignment decisions. vrAIn, which is implemented using an open-source LTE stack over a Linux platform, can achieve substantial savings in the used CPU resources while maintaining the target QoS for the attached terminals and maximizing throughput when there is a deficit of computational capacity., The work of University Carlos III of Madrid was supported by H2020 5G-MoNArch project (grant agreement no. 761445) and H2020 5G-TOURS project (grant agreement no. 856950). The work of NEC Laboratories Europe was supported by H2020 5G-TRANSFORMER project (grant agreement no. 761536) and 5GROWTH project (grant agreement no. 856709). The work of University of Cartagena was supported by Grant AEI/FEDER TEC2016-76465-C2-1-R (AIM) and Grant FPU14/03701.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

On slice isolation options in the transport network and associated feasibility indicators

  • Contreras Murillo, Luis Miguel
  • Ordoñez Lucena, Jose Antonio
Proceedings of: IEEE 7th International Conference on Network Softwarization (NetSoft 2021), Isolation is one of the more relevant attributes associated to the idea of network slicing, introduced by 5G services. Through isolation it is expected that slices from different customers could gracefully coexist without interfering each other, in the sense that whatever misbehavior or unforeseen demand from one slice customer could not affect the communication service received by any other slice customer supported atop the same physical transport infrastructure. This paper surveys and compare different technical approaches that can be taken for providing distinct isolation levels in the transport network, as a major component of end-to-end network slices. Furthermore, a number of isolation feasibility indicators are defined and proposed. These indicators are based on the approaches referred before, as a mean of guiding orchestration decisions at the time of provisioning or reconfiguring the transport slices in the network., This work has been partly funded by the 5G-PPP projects 5G-DIVE (Grant Agreement no. 859881), 5GROWTH (G. A. no. 856709) and 5G-VINNI (G. A. no. 815279).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

Orchestrating Edge- and Cloud-based Predictive Analytics Services

  • Chintapalli, Venkatarami Reddy
  • Kondepu, Koteswararao
  • Sgambelluri, Andrea
  • Franklin A., Antony
  • Tamma, Bheemarjuna Reddy
  • Castoldi, Piero
  • Valcarenghi, Luca
This paper has been presented at the 2020 European Conference on Networks and Communications (EuCNC), In the Zero-touch network and service management (ZSM) architecture, devised by ETSI, making predictions on the observed data is among the functions provided by the analytics block of the control loop cycle. Prediction performance depends on several parameters, such as the utilized computational resources, the leveraged prediction techniques, the deployment location of the prediction tools with respect to the data.This paper proposes a Hybrid Forecast Framework (HFF) running at the network edge or in the cloud to provide fore-casting with the performance required by the control loop cycle. Forecasting at the edge might shorten the control loop cycle if resources shall be made available locally where data is collected. However, in general, edge computational resources are less abundant than the cloud ones, thus causing longer time to perform the prediction. On the opposite, forecasting in the cloud might require more time for the data to reach the utilized tools but more computational resources could be exploited. The HFF is based on utilizing traditional time series analysis prediction algorithms to minimize the utilized resources and energy at the edge while it exploits AI/ML tools to make predictions in the cloud.Results show that for short lead time (i.e., the time, in the future, at which the status of the considered parameter is predicted) edge-based prediction exploiting time series analysis provides better accuracy, requires less resources and time (thus energy) than cloud-based prediction. However, if the lead time is long, cloud-based prediction exploiting Artificial Intelligence/Machine Learning (AI/ML) provides better accuracy. Thus, if the lead time is long, it is preferable because the long lead time compensates for the higher time for prediction, mainly due to data transfer., This work has been partially supported by the project "Scheme for Promotion of Academic and Research Collaboration (SPARC)", MHRD, Govt. of India and the EU Commission through the 5GROWTH project (grant agreement no. 856709).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

AZTEC: anticipatory capacity allocation for zero-touch network slicing

  • Bega, Dario
  • Gramaglia, Marco
  • Fiore, Marco
  • Banchs Roca, Albert
  • Costa-Pérez, Xavier
Proceeding of: IEEE INFOCOM 2020 - IEEE Conference on Computer Communications 6-9 July 2020 (Virtual Conference), The combination of network softwarization with network slicing enables the provisioning of very diverse services over the same network infrastructure. However, it also creates a complex environment where the orchestration of network resources cannot be guided by traditional, human-in-the-loop network management approaches. New solutions that perform these tasks automatically and in advance are needed, paving the way to zero-touch network slicing.In this paper, we propose AZTEC, a data-driven framework that effectively allocates capacity to individual slices by adopting an original multi-timescale forecasting model. Hinging on a combination of Deep Learning architectures and a traditional optimization algorithm, AZTEC anticipates resource assignments that minimize the comprehensive management costs induced by resource overprovisioning, instantiation and reconfiguration, as well as by denied traffic demands.Experiments with real-world mobile data traffic show that AZTEC dynamically adapts to traffic fluctuations, and largely outperforms state-of-the-art solutions for network resource orchestration., The work of University Carlos III of Madrid was supported by H2020 5G-TOURS project (grant agreement no. 856950). The work of NEC Laboratories Europe was supported by H2020 5GROWTH project (grant agreement no. 856709). The research of M. Fiore was partially supported by ANR CANCAN project (ANR-18-CE25-0011).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

Cloud native computing for Industry 4.0: Challenges and opportunities

  • Gil, Guilherme
  • Corujo, Daniel
  • Pedreiras, Paulo
Proceedings of: 26th IEEE International Conference on Emerging Technologies and Factory Automation (ETFA ), 7-10 Sept. 2021, Vasteras, Sweden., Cloud-based architectures are advantageous in aspects such as scalability, reliability and resource utilization efficiency, to name just a few, thus being considered one of the pillars of Industry 4.0. However, in this domain, cloud computing platforms are subject to specific requirements, namely in what concerns real-time performance, determinism and fault-tolerance. This paper focuses on cloud native computing, which is an emerging and promising cloud-computing paradigm, specifically addressing its applicability to real-time systems. Firstly, it introduces the architecture of cloud native applications, discussing their principles, potential advantages and challenges. Then it addresses the opportunities and constraints of such technologies when applied to industrial real-time systems., This work has been supported by EC H2020 5GPPP 5Growth project (Grant 856709).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

COTORRA: Context-Aware testbed for robotic applications

  • Groshev, Milan
  • Martín Pérez, Jorge
  • Antevski, Kiril
  • Oliva Delgado, Antonio de la
  • Bernardos Cano, Carlos Jesús
Proceedings of: 1st Workshop on Serverless mobile networking for 6G Communications, 25 June 2021, Virtual conference, Edge computing have received considerable attention as a promising candidate for the evolution of robotic systems. In this work, we propose COTORRA, an Edge driven robotic testbed that combines context information with robot sensor data to validate innovative concepts for robotic systems prior to being applied in a production environment. We have tested COTORRA in a controlled university environment as an easy applicable, serverless, and modular testbed on top of commodity network infrastructure. COTORRA supports pluggable robotic applications. To verify its feasibility and assess its performance, we ran a set of experiments that show how autonomous navigation applications can achieve target latencies bellow 15 ms, and perform an inter-domain Distributed Ledger Technology (DLT) federation within 19 seconds., This work has been partially funded by the EU H2020 5GROWTH Project (grant no. 856709) and by the H2020 collaborative Europe/Taiwan research project 5G-DIVE (grant no. 859881).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

Efficient restoration of simultaneous transport services within an NFV infrastructure

  • Martínez, Ricardo
  • Vettori, Luca
  • Baranda, Jorge
  • Mangues-Bafalluy, Josep
  • Zeydan, Engin
Proceedings of: IEEE Global Communications Conference (GLOBECOM), 7-11 Dec. 2021, Madrid, Spain., In 5G networks, heterogeneous vertical services with different requirements are rolled out over a common multi-technology infrastructure. A resource orchestrator entity automatically coordinates the operations and functions to support the service's lifecycle management (i.e., creation, update and termination). Moreover, it is essential that service needs are continuously assured even if transport network anomalies (e.g., link failures) occur. Herein, we present an implemented resource orchestrator architecture integrating monitoring capabilities to attain closed-loop operations for: i) gathering monitored information; ii) detecting transport network anomalies; and iii) triggering the required action (e.g., restoration) to keep the service continuity. When a link failure happens, several transport services may be disrupted requiring to be immediately restored. To this end, we propose a novel on-line restoration algorithm called as Global Concurrent Optimization (GCO). The GCO algorithm aims at attaining an enhanced restorability performance compared to a more traditional restoration algorithm (referred to as 1-by-1). Both algorithms are experimentally compared on top of the deployed resource orchestrator architecture. The evaluation is done upon both dynamic service arrival/departure and link failure generation using different performance metrics: the average restorability, the average network resource utilization, and the restoration computational time., Work supported in part by EU Commission H2020 5Growth project (Grant No. 856709), Spanish MICINN AURORAS (RTI2018-099178-B-I00) and Spanish MINECO 5G-REFINE (TEC2017-88373-R) projects and Generalitat de Catalunya grant 2017 SGR 1195.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

Demo: AIML-as-a-service for SLA management of a digital twin virtual network service

  • Baranda, Jorge
  • Zeydan, Engin
  • Casetti, C.
  • Chiasserini, C. F.
  • Malinverno, Marco
  • Puligheddu, C.
  • Groshev, Milan
  • Magalhaes Guimaraes, Carlos Eduardo
  • Tomakh, K.
  • Kucherenko, D.
  • Kolodiaznhyi, O.
  • Mangues-Bafalluy, Josep
Proceedings of: IEEE INFOCOM 2021 - IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS)., This demonstration presents an AI/ML platform that is offered as a service (AIMLaaS) and integrated in the management and orchestration (MANO) workflow defined in the project 5Growth following the recommendations of various standardization organizations. In such a system, SLA management decisions (scaling, in this demo) are taken at runtime by AI/ML models that are requested and downloaded by the MANO stack from the AI/ML platform at instantiation time, according to the service definition. Relevant metrics to be injected into the model are also automatically configured so that they are collected, ingested, and consumed along the deployed data engineering pipeline. The use case to which it is applied is a digital twin service, whose control and motion planning function has stringent latency constraints (directly linked to its CPU consumption), eventually determining the need for scaling out/in to fulfill the SLA., Work supported in part by EU Commission H2020 5Growth project (Grant No. 856709) and H2020 Europe/Taiwan 5G-Dive project (Grant No. 859881).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

PI2 for P4: An Active Queue Management Scheme for Programmable Data Planes

  • Papagianni, Chrysa
  • Schepper, Koen De
This paper has been presented at CoNEXT 2019: The 15th International Conference on emerging Networking EXperiments and Technologies, Network programmability brings agility, flexibility and rapid introduction of new capabilities, towards supporting application requirements. Enabling programmable traffic management at the data plane can provide significant benefits pertaining to Quality of Service (QoS). To this end we adopt P4, domain specific language for programming network switches, to implement an Active Queue Management (AQM) scheme for reducing queuing delay. We verify the design with a proof of concept implementation of PI2, a modern AQM that requires only basic bit manipulations at the data plane., This work was supported by H2020 5GROWTH project (grant agreement no. 856709)
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

DLT federation for Edge robotics

  • Antevski, Kiril
  • Groshev, Milan
  • Baldoni, Gabriele
  • Bernardos Cano, Carlos Jesús
The concept of federation in 5G and NFV networks aims to provide orchestration of services across multiple administrative domains. Edge robotics, as a field of robotics, implements the robot control on the network edge by relying on low-latency and reliable access connectivity. In this paper, we propose a solution that enables Edge robotics service to expand its service footprint or access coverage over multiple administrative domains. We propose application of Distributed ledger technologies (DLTs) for the federation procedures to enable private, secure and trusty interactions between undisclosed administrative domains. The solution is applied on a real-case Edge robotics experimental scenario. The results show that it takes around 19 seconds to deploy & federate a Edge robotics service in an external/anonymous domain without any service down-time., This work has been partially funded by the EU H2020 5GROWTH Project (grant no. 856709) and by the H2020 collaborative Europe/Taiwan research project 5G-DIVE (grant no. 859881).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

Remote Control of a Robot Rover Combining 5G, AI, and GPU Image Processing at the Edge

  • Civerchia, F.
  • Giannone, F.
  • Kondepu, K.
  • Castoldi, P.
  • Valcarenghi, L.
  • Bragagnini, A.
  • Gatti, F.
  • Napolitano, A.
  • Borromeo, Justine Cris
This paper has been presented at 2020 Optical Fiber Communications Conference and Exhibition (OFC), The demo shows the effectiveness of a low latency remote control based on 5G and image processing at the edge exploiting artificial intelligence and GPUs to make a robot rover slalom between posts., This work has been partially supported by TIM under the Cooperation Agreement with Scuola Superiore Sant’Anna for the 5G MISE Trial in Bari and Matera 2018-2022 and the EU Commission through the 5GROWTH project (grant agreement no. 856709).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

On the Integration of AI/ML-based scaling operations in the 5Growth platform

  • Baranda, Jorge
  • Mangues-Bafalluy, Josep
  • Zeydan, Engin
  • Vettori, L.
  • Martínez, Ricardo
  • Li, Xi
  • García Saavedra, Andrés
  • Chiasserini, C. F.
  • Casetti, C.
  • Tomakh, K.
  • Kolodiazhnyi, O.
  • Bernardos Cano, Carlos Jesús
This paper has been presented at 2020 IEEE Conference on Network Function Virtualization and Software Defined Networks, The automated assurance of vertical service level agreements (SLA) is a challenge in 5G networks. The EU 5Growth project designs and develops a 5G End-to-End service platform that integrates Artificial Intelligence (AI) and Machine Learning (ML) techniques for any decision-making process in the management and orchestration (MANO) stack. This paper presents the detailed architecture and first prototype of the 5Growth platform taking AI/ML-based network service auto-scaling decisions. This also includes the modification of the ETSI network service descriptors for requesting AI/ML-based decisions for orchestration problems and the integration of a data engineering pipeline for real-time data gathering and model execution. Our evaluation shows that AI/ML-related service handling operations (1&-2 s.) are well below instantiation/termination procedures (80/60 s., respectively). Furthermore, online classification can be performed in the order of hundreds of milliseconds (600 ms)., This work has been partially funded by the EU H2020 5Growth Project (grant no. 856709), by MINECO grant TEC2017-88373-R (5G-REFINE) and Generalitat de Catalunya grant 2017 SGR 1195.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

OKpi: All-KPI Network Slicing Through Efficient Resource Allocation

  • Martín Pérez, Jorge
  • Malandrino, Francesco
  • Chiasserini, C. F.
  • Bernardos Cano, Carlos Jesús
This paper has been presented at INFOCOM 2020: IEEE International Conference on Computer Communications, Networks can now process data as well as transporting it; it follows that they can support multiple services, each requiring different key performance indicators (KPIs). Because of the former, it is critical to efficiently allocate network and computing resources to provide the required services, and, because of the latter, such decisions must jointly consider all KPIs targeted by a service. Accounting for newly introduced KPIs (e.g., availability and reliability) requires tailored models and solution strategies, and has been conspicuously neglected by existing works, which are instead built around traditional metrics like throughput and latency. We fill this gap by presenting a novel methodology and resource allocation scheme, named OKpi, which enables high-quality selection of radio points of access as well as VNF (Virtual Network Function) placement and data routing, with polynomial computational complexity. OKpi accounts for all relevant KPIs required by each service, and for any available resource from the fog to the cloud. We prove several important properties of OKpi and evaluate its performance in two real-world scenarios, finding it to closely match the optimum., This work was supported by the EU Commission through the 5GROWTH project (grant agreement no. 856709).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

OpenFlowMon: a fully distributed monitoring framework for virtualized environments

  • Cobos Dominguez, Antonio
  • Magalhaes Guimaraes, Carlos Eduardo
  • Oliva Delgado, Antonio de la
  • Zabala Orive, Aitor
Proceedings of: 2021 IEEE Conference on Network Function Virtualization and Software Defined Networks (NFV-SDN), 9 November 2021, Heraklion, Greece., Network monitoring allows a continuous assessment on the health and performance of the network infrastructure. With the significant change on how networks are deployed and operated, mainly due to the advent of virtualization technologies, alternative monitoring approaches are emerging to provide a finer-grained flow monitoring to complement already existing mechanisms and capabilities. In this paper, we proposed and developed an Open-Source Flow Monitoring Framework (OpenFlowMon), a fully distributed monitoring framework implemented solely with open-source solutions. This framework is used to assess the performance and the overhead introduced by two different flow monitoring approaches: (i) switch level and (ii) compute node level monitoring. Results show that monitoring at compute node level not only reduces the overhead but also mitigates a potential complex post-processing in east-to-west traffic., This work has been (partially) funded by H2020 EU/TW 5G-DIVE (Grant 859881) and H2020 5Growth (Grant 856709).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

SliMANO: An Expandable Framework for the Management and Orchestration of End-to-end Network Slices

  • Meneses, Flávio
  • Fernandes, Manuel
  • Corujo, Daniel
  • Aguiar, Rui L.
This paper has been presented at The IEEE 8th International Conference on Cloud Networking (CloudNet), This paper proposes a slice management and orchestration framework for abstracting the instantiation of end-to-end network slices, which are composed by a chain of both physical and virtual network functions. In this line, the proposed SliMANO framework is a plug-in based system that requests network resources and coordinates the interaction among network orchestration entities for its instantiation and chaining in order to perform an end-to-end slice. These entities could range from management and orchestration (MANO), Software Defined Networking (SDN) controllers and Radio Access Network (RAN) controllers. A proof-of-concept prototype was implemented and experimentally evaluated, with results showcasing its feasibility. The results revealed a increase in the delay, associated with instantiation and deletion operations, when compared with the recently introduced network slicing feature (NetSlice) of the Open-source Management and Orchestration (OSM)., This work is funded by FCT/MEC through national funds under the project PTDC/EEI-TEL/30685/2017 and by the European Commission through the H2020 project 5GROWTH (Project ID 856709).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

5Growth data-driven AI-based scaling

  • De Vleeschauwer, Danny
  • Baranda, Jorge
  • Mangues-Bafalluy, Josep
  • Chiasserini, Carla Fabiana
  • Malinverno, Marco
  • Puligheddu, Corrado
  • Magoula, Lina
  • Martín Pérez, Jorge
  • Barmpounakis, Sokratis
  • Kondepu, Koteswararao
  • Valcarenzhi, Luca
  • Li, Xi
  • Papagianni, Chrysa
  • García Saavedra, Andrés
Proceedings of: Joint European Conference on Networks and Communications & 6G Summit (EuCNC/6G Summit), 8-11 June 2021, Porto, Portugal., This paper presents a data-driven approach leveraging AI/ML models to automate the service scaling operation and, in this way, meet the service requirements while minimizing the consumption of network, computing, and storage resources. This approach is integrated into the 5Growth service management software platform. In particular, a prototype was developed to demonstrate how the novel 5Growth AI/ML platform can be used in a closed-loop automation system to support the automated service scaling operation. Furthermore, a number of additional ML-based approaches are developed in the context of eMBB and C-V2N scenarios, which can be embedded into the system for handling more complex use cases., This work has been partially supported by EC H2020 5GPPP 5Growth project (Grant 856709).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

NFV Service Federation: enabling Multi-Provider eHealth Emergency Services

  • Baranda, Jorge
  • Mangues-Bafalluy, Josep
  • Vettori, L.
  • Martínez, R.
  • Antevski, Kiril
  • Girletti, Luigi
  • Bernardos Cano, Carlos Jesús
  • Tomakh, K.
  • Kucherenko, D.
  • Landi, Giada
  • Brenes, Juan
  • Li, Xi
  • Costa-Pérez, Xavier
  • Ubaldi, F.
  • Imbarlina, G.
  • Gharbaoui, M.
One of the key challenges in developing 5G/6G is to offer improved vertical service support providing enlarged service flexibility, coverage and connectivity while enhancing the business relations among different stakeholders. To address this challenge, Network Service Federation (NSF) is a required feature to enable the deployment and the management of vertical services that may span multiple provider domains owned by different operators and/or service providers. In this demonstration, we show our proposed NSF solution to dynamically deploy an eHealth network service across multiple provider domains at different locations., Work supported in part by EU Commission H2020 5G-TRANSFORMER project (Grant No. 761536), H2020 5Growth project (Grant No. 856709), Spanish MINECO grant TEC2017-88373-R (5G-REFINE) and Generalitat de Catalunya grant 2017 SGR 1195.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

Demo: Nuberu - A Reliable DU DesignSuitable for Virtualization Platforms

  • García Avilés, Ginés
  • García Saavedra, Andrés
  • Gramaglia, Marco
  • Costa-Pérez, Xavier
  • Serrano Yáñez-Mingot, Pablo
  • Banchs Roca, Albert
We demonstrate Nuberu. The scenario consists of a DU under test (DuT), and one or more DUs sharing computing resources. A dashboard lets us control the type of DuT: Baseline, implemented with vanilla srsRAN, or Nuberu; the number of competing vDUs; and their SNR. A second screen shows real-time metrics: the processing latency of the TBs from each vDU instance; the throughput performance of DuT; the processing latency of DU jobs from DuT; and the ratio of latency constraint violations of DuT jobs. We show how the throughput attained by the baseline DU approach collapses upon sufficiently high computing interference from the competing DUs. Conversely, we show that the DU design introduced in [3] preserves reliability irrespective of the computing interference., This work has been supported by the EC through Grant No. 101017109 (DAEMON) and Grant No. 856709 (5Growth).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

Serving HTC and critical MTC in a RAN slice

  • Mancuso, Vincenzo
  • Castagno, Paolo
  • Sereno, Matteo
  • Marsan, Marco Ajmone
Proceedings of: IEEE 22nd International Symposium on a World of Wireless, Mobile and Multimedia Networks (WoWMoM), 7-11 June 2021, Pisa, Italy., We consider a slice of a radio access network where human and machine users access services with either high throughput or low latency requirements. The slice offers both eMBB and URLLC service categories to serve HTC (Human-Type Communication) and MTC (Machine-Type Communication) traffic. We propose to use eMBB for both HTC and MTC, transferring machine traffic to URLLC only when eMBB is not able to meet the low latency requirements of MTC. We show that by so doing the slice is capable of providing very good performance to about one hundred MTC users under high HTC traffic conditions. Instead, running time-critical MTC over only eMBB is not doable at all, whereas using URLLC suffices for at most a few tens of devices. Therefore, our approach improves the number of users served by the slice by one order of magnitude, without requiring extra resources or compromising performance. To study system performance we develop a novel analytical model of uplink packet transmissions, which covers both legacy eMBB-or URLLC-based MTC, as well as our compound approach. Our model allows to tune slice parameters so as to achieve the desired balance between HTC and MTC service guarantees. We validate the model against detailed simulations using as an example an autonomous driving scenario., V. Mancuso was supported by the Ramon y Cajal grant RYC-2014-16285 from the Spanish Ministry of Economy and Competitiveness. This work was partially supported by the EU 5GROWTH project (Grant No. 856709), and by the Region of Madrid through the TAPIR-CM project (S2018/TCS-4496).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

Arbitrating Network Services in 5G Networks for Automotive Vertical Industry

  • Baranda, Jorge
  • Vettori, L.
  • Martínez, R.
  • Avino, G.
  • Chiasserini, C. F.
  • Puligheddu, C.
  • Casetti, C.
  • Brenes, Juan
  • Landi, Giada
  • Kondepu, K.
  • Paolucci, F.
  • Fichera, S.
  • Valcarenghi, L.
  • Mangues-Bafalluy, Josep
This paper has been presented at IEEE INFOCOM 2020 - IEEE Conference on Computer Communications Workshops, This demonstration shows how the 5G-TRANSFORMER platform, and more specifically the vertical slicer, is capable of arbitrating vertical services. In this context, arbitration refers to handling the various services of a given vertical customer according to their SLA requirements, service priorities, and resource budget available to the vertical. In this demo, a low priority video service of the automotive vertical is terminated when a high-priority intersection collision avoidance service needs to be instantiated and there are not enough resources allowing all services to be run in parallel. All these services are deployed with the help of the 5G-TRANSFORMER platform in a multi-PoP scenario, with PoPs in Barcelona (Spain), Turin (Italy), and Pisa (Italy), featuring a high variety of transport and computing technologies., Work supported in part by EU Commission H2020 5G-TRANSFORMER project (Grant No. 761536), H2020 5Growth project (Grant No. 856709), Spanish MINECO grant TEC2017-88373-R (5G-REFINE) and Generalitat de Catalunya grant 2017 SGR 1195.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

Efficiency gains due to network function sharing in CDN-as-a-Service slicing scenarios

  • Contreras, Luis M.
  • Solano, Alberto
  • Cano, Francisco
  • Folgueira, Jesus
Proceedings of: IEEE 7th International Conference on Network Softwarization (NetSoft), 28 June-2 July 2021, Tokyo, Japan., The consumption of video contents is currently dominating the traffic observed in ISP networks. The distribution of that content is usually performed leveraging on CDN caches storing and delivering multimedia. The advent of virtualization is bringing attention to the CDN as use case for virtualizing the cache function. In parallel, there is a trend on sharing network infrastructures as a way of reducing deployment costs by ISPs. Then, an interesting scenario emerges when considering the possibility of sharing virtualized cache functions among ISPs sharing a common physical infrastructure, mostly considering that usually those ISPs offer similar content catalogues to final end users. This paper investigates through simulations the potential efficiencies that can be achieved when sharing a virtual cache function if compared to the classical approach of independent virtual caches operated per ISP., This work has been partly funded by the project 5GROWTH (Grant Agreement no. 856709).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

Information Exchange to Support Multi-Domain Slice Service Provision for 5G/NFV

  • Solano, Alberto
  • Contreras Murillo, Luis Miguel
This paper has been presented at 2020 IFIP Networking Conference (Networking), Service providers will highly rely on network softwarization operational trends for addressing 5G verticals’ market, leveraging on multi-domain cloud computing capabilities offered by different stakeholders. Consequently, multi-domain scenarios in which diverse providers form a federation will become more relevant for 5G and beyond. Due to the intrinsic limited footprint and infrastructure capacity offered by just a single provider, the progression towards these scenarios will require mechanisms to support an efficient orchestration of network slices over multi-provider assets. In this paper, we analyse the convenience of enabling timely exchange of information in terms of guaranteed latency and resource availability among various telecommunication providers participating in a federation to effectively ensure the deployment of 5G vertical services and reduce the service provision blocking probability., This work has been partly funded by the project 5GROWTH (Grant Agreement no. 856709).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

A Q-learning strategy for federation of 5G services

  • Antevski, Kiril
  • Martín Pérez, Jorge
  • García Saavedra, Andrés
  • Bernardos Cano, Carlos Jesús
  • Li, Xi
  • Baranda, Jorge
  • Mangues-Bafalluy, Josep
  • Martínez, Ricardo
  • Vettori, Luca
This paper has been presented at the 2020 IEEE International Conference on Communications, 5G networks aim to provide orchestration of services across multiple administrative domains through the concept of federation. In this paper, we are exploring the federation feature of a platform for 5G transport network of vertical services. Then we formulate the decision problem that directly impacts the revenue of 5G administrative domains, and we propose as solution a Q-learning algorithm. The simulation results show near optimum profit maximization and a well-trained Q-learning algorithm can outperform the intuitive "greedy" approach in a realistic scenario., This work has been partially funded by the EC H2020 5G-TRANSFORMER Project (grant no. 761536) and the EU H2020 5GROWTH Project (grant no. 856709).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

An RL approach to radio resource management in heterogeneous virtual RANs

  • Tripathi, Sharda
  • Puligheddu, Corrado
  • Chiasserini, Carla Fabiana
Proceedings of: 16th Annual Conference on Wireless On-demand Network Systems and Services Conference (WONS), 9-11 March 2021, Klosters, Switzerland., 5G networks are primarily designed to support a wide range of services characterized by diverse key performance indicators (KPIs). A fundamental component of 5G networks, and a pivotal factor to the fulfillment of the services KPIs, is the virtual radio access network (RAN), which allows high flexibility on the control of the radio link. However, to fully exploit the potentiality of virtual RANs in non-stationary environments, an efficient mapping of the rapidly varying context to radio control decisions is not only essential, but also challenging owing to the non-trivial interdependence of network and channel conditions. In this paper, we propose CAREM, an RL framework for dynamic radio resource allocation, which selects the best link and modulation and coding scheme (MCS) for packet transmission, so as to meet the KPI requirements in heterogeneous virtual RANs. To show its effectiveness in real-world conditions, we provide a proof-of-concept through actual testbed implementation. Experimental results demonstrate that CAREM enables an efficient radio resource allocation, for any of the considered time periodicity of the decision-making process., This work has been supported by the EC H2020 5GPPP 5GROWTH project (Grant No. 856709.)
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

R-Learning-based admission control for service federation in multi-domain 5G networks

  • Bakhshi, Bahador
  • Mangues-Bafalluy, Josep
  • Baranda, Jorge
Proceedings of: IEEE Global Communications Conference (GLOBECOM), 7-11 Dec. 2021, Madrid, Spain., Network service federation in 5G/B5G networks enables service providers to extend service offering by collaborating with peering providers. Realizing this vision requires interoperability among providers towards end-to-end service orchestration across multiple administrative domains. Smart admission control is fundamental to make such extended offering profitable. Without prior knowledge of service requests, the admission controller (AC) either determines the domain to deploy each demand or rejects it to maximize the long-term average profit. In this paper, we first obtain the optimal AC policy by formulating the problem as a Markov decision process, which is solved through the policy iteration method. This provides the theoretical performance bound under the assumption of known arrival and departure rates of demands. Then, for practical solutions to be deployed in real systems, where the rates are not known, we apply the Q-Learning and R-Learning algorithms to approximate the optimal policy. The extensive simulation results show that learning approaches outperform the greedy policy and are capable of getting close to optimal performance. More specifically, R-learning always outperformed the rest of practical solutions and achieved an optimality gap of 3-5% independent of the system configuration, while Q-Learning showed lower performance and depended on discount factor tuning., This work has been partially funded by the MINECO grant TEC2017-88373-R (5G-REFINE), the EC H2020 5Growth Project (grant no. 856709), and Generalitat de Catalunya grant 2017 SGR 1195.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

An RL approach to radio resource management in heterogeneous virtual RANs

  • Tripathi, Sharda
  • Puligheddu, Corrado
  • Chiasserini, Carla Fabiana
Proceedings of: 2021 16th Annual Conference on Wireless On-demand Network Systems and Services Conference (WONS), 9-11 March 2021, Klosters, Switzerland., 5G networks are primarily designed to support a wide range of services characterized by diverse key performance indicators (KPIs). A fundamental component of 5G networks, and a pivotal factor to the fulfillment of the services KPIs, is the virtual radio access network (RAN), which allows high flexibility on the control of the radio link. However, to fully exploit the potentiality of virtual RANs in non-stationary environments, an efficient mapping of the rapidly varying context to radio control decisions is not only essential, but also challenging owing to the non-trivial interdependence of network and channel conditions. In this paper, we propose CAREM, an RL framework for dynamic radio resource allocation, which selects the best link and modulation and coding scheme (MCS) for packet transmission, so as to meet the KPI requirements in heterogeneous virtual RANs. To show its effectiveness in real-world conditions, we provide a proof-of-concept through actual testbed implementation. Experimental results demonstrate that CAREM enables an efficient radio resource allocation, for any of the considered time periodicity of the decision-making process., This work has been supported by the EC H2020 5GPPP 5GROWTH project (Grant No. 856709.)
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

5 Growth: Secure and reliable network slicing for verticals

  • Cunha, Vitor A.
  • Maroulis, Nikolaos
  • Papagianni, Chrysa
  • Sacido, Javier
  • Jiménez, Manuel
  • Ubaldi, Fabio
  • Gharbaoui, Molka
  • Chang, Chia-Yu
  • Koursioumpas, Nikolaos
  • Tomakhxi, Konstantin
  • Corujo, Daniel
  • Barraca, Joao P.
  • Barmpounakis, Sokratis
  • Kucherenkoxi, Denys
  • Giorgetti, Alessio
  • Boddi, Andrea
  • Valcarenghi, Luca
  • Kolodiazhnyixi, Oleksii
  • Zabala Orive, Aitor
  • Salvat, Josep Xavier
  • Garcia-Saavedra, Andres
Proceedings of: Joint European Conference on Networks and Communications & 6G Summit (EuCNC/6G Summit), 8-11 June 2021, Porto, Portugal., Network slicing is well-recognized as a core 5G technology to enable heterogeneous vertical services sharing the same infrastructure. In this context, the H2020 5Growth project extends baseline 5G management and orchestration platforms to manage the life-cycle of real end-to-end, reliable, and secure network slices with performance guarantees. In this paper, we present 5Growth's approaches to (i) attain isolation across network slices, (ii) provide secure interfaces towards third parties, and (iii) exploit AI/ML to achieve reliability through automated anomaly detection. In our quest towards validating full-fledged 5G pilots, we demonstrate our slicing mechanisms in PoCs that include interacting with ICT-17 infrastructure., This work has been partially supported by EC H2020 5GPPP 5Growth project (Grant 856709).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

Handover Prediction Integrated with Service Migration in 5G Systems

  • Abdah, Hadeel
  • Barraca, João Paulo
  • Aguiar, Rui L.
This paper has been presented at ICC 2020 - 2020 IEEE International Conference on Communications (ICC), As the research community inclines toward adopting increasingly complex techniques for future networks, and simple methods are often ignored, being labeled as trivial. In this paper, we argue that simple methods can sometimes outperform more sophisticated ones. We demonstrate that by evaluating two prediction mechanisms to forecast mobile user’s handovers exploiting user-network association patterns. We perform a series of experiments on real-world data, evaluating the performance characteristics of such methods over more sophisticated and complex prediction techniques. Furthermore, we discuss how to easily bootstrap these mechanisms into the 5G network architecture. We suggest the use of these methods associated with Multi-access Edge Computing (MEC) scenarios, as a mean to identify favorable edge nodes to host the mobile applications, to best provide continuous and QoS-aware service for mobile users., This work is funded by FCT/MEC through national funds and when applicable co-funded by FEDER - PT2020 partnership agreement under the project UID/EEA/50008/2019, Fundacao para a Ciencia e Tecnologia under grant SFRH/BD/136361/2018, and by the European Commission through the H2020 project 5GROWTH (Project ID 856709).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

The case for serverless mobile networking

  • Gramaglia, Marco
  • Serrano Yáñez-Mingot, Pablo
  • Banchs Roca, Albert
  • García Avilés, Ginés
  • García-Saavedra, Andrés
  • Pérez, Ramón
Proceeding of: IFIP Networking Conference 2020, Paris, France, 22-26 June 2020 (Virtual conference)., The softwarization of communication networks provides notable benefits, such as flexibility, improved resource efficiency, and commoditization. In exchange, softwarization requires an increased management overhead and the need to re-design network operation. While the mobile networking eco-system is currently adapting this new paradigm with other network-related aspects (e.g., network slicing), cloud computing already addressed such problems with the introduction of serverless architectures, also known as Function as a Service (FaaS). With this approach, the software is decomposed into its minimum building blocks, i.e., functions, maximizing scalability, resource efficiency, and flexibility. In this paper, we analyze the potential adoption of the FaaS paradigm by the mobile networking ecosystem, discussing the implicit advantages, the challenges to address, and some solutions to overcome them., This work was partially supported by the European Commission in the framework of the H2020 5G-PPP 5G-EVE 5GROWTH, and 5G-TOURS projects.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

Dynamic Interdomain Network Slicing for verticals in the 5Growth project

  • Fonseca, João
  • Alegria, João
  • Cunha, Vitor A.
  • Quevedo, José
  • Santos, David
  • Gomes, Diogo
  • Barraca, João P.
  • Corujo, Daniel
  • Aguiar, Rui L.
Proceedings of: IEEE Conference on Network Function Virtualization and Software Defined Networks (NFV-SDN), 9-11 Nov. 2021, Heraklion, Greece., This paper proposes and validates a Interdomain Network Slicing framework for verticals, allowing them to directly participate in the establishment and control of end-to-end Communication Services deployment across multiple inter-operator domains. The framework progresses the means made available by different standards and research initiatives to enhance service requesting and provisioning interfaces for the stakeholders involved, namely operators and verticals. The framework is validated under two different use cases, showcasing effective end-to-end service instantiation and a first assessment towards dynamic service modification capability., This work has been supported by EC H2020 5GPPP 5Growth project (Grant 856709).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

Demo: Scaling Composite NFV-Network Services

  • Baranda, Jorge
  • Mangues-Bafalluy, Josep
  • Vettori, Luca
  • Martínez, Ricardo
This poster has been presented at Mobihoc '20: The Twenty-first ACM International Symposium on Theory, Algorithmic Foundations, and Protocol Design for Mobile Networks and Mobile Computing, The composition of NFV-network services (NSs) helps 5G networks to provide the required flexibility and dynamicity to satisfy the needs of vertical industries. The mentioned capabilities are required not only at instantiation time but also during operation time to maintain the corresponding service level agreements in the pres-ence of changing network conditions. In this demonstration, we present the capabilities of the 5Growth platform to handle the scal-ing of composite NSs under different situations. In particular, we show the scaling of a composite NS implying the update of the required deployed nested NS components and its interconnections but also how the scaling of a shared single NS entails updates on its associated composite NSs., This work has been partially funded by the EC H2020 5G-Growth Project (grant no. 856709), by MINECO grant TEC2017-88373-R (5G-REFINE) and Generalitat de Catalunya grant 2017 SGR 1195.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

Experimental validation of compute and network resource abstraction and allocation mechanisms within an NFV infrastructure

  • Martínez, Ricardo
  • Vettori, Luca
  • Baranda, Jorge
  • Mangues-Bafalluy, Josep
  • Zeydan, Engin
Proceedings of: IFIP/IEEE International Symposium on Integrated Network Management (IM), 17-21 May 2021, Bordeaux, France., 5G supported capabilities (e.g., slicing) enable accommodating heterogeneous vertical services having their own requirements over a common cloud and transport infrastructure. In this context, the EU-H2020 5Growth project defines a service and infrastructure orchestration architecture to automatically deploy network services (NSes) fulfilling vertical demands. In this architecture, the Service Orchestrator (5Gr-SO), as a service provider, maps the vertical service needs into NS requirements (e.g., CPU, RAM, bandwidth, etc.). The 5Gr-SO interacts with an underlying infrastructure orchestrator referred to as 5Gr-RL. The 5Gr-RL, as an infrastructure provider, handles two main functions: i) abstraction of the resources exposed to the 5GrSO, and ii) fine-grained resource selection. Different interaction forms between both 5Gr-SO and 5Gr-RL arise differing in the exchanged abstracted information and resource allocation. We present two 5Gr-SO and 5Gr-RL interaction solutions stemming from two 5Gr-RL operational modes: Infrastructure Abstraction (InA) and Connectivity Service Abstraction (CSA). In the InA approach, the 5Gr-SO is granted with an aggregated view of the computing resources and a set of transport logical links between the cloud locations. In the CSA strategy, besides the aggregated view of the cloud resources, the logical links are associated to potential connectivity service types. Both InA and CSA strategies are presented describing their pros and cons. Moreover, the designed workflows (involving the devised abstraction and allocation algorithms) between the 5Gr-SO and 5Gr-RL entities are experimentally validated. Scalability studies are conducted upon two different cloud and transport infrastructure sizes in terms of the abstraction composition time, the expansion computation time, and total NS deployment time., Work supported in part by EU Commission H2020 5Growth project (Grant No. 856709), Spanish MICINN AURORAS (RTI2018-099178-B-I00) and Spanish MINECO 5G-REFINE (TEC2017-88373-R) projects and Generalitat de Catalunya grant 2017 SGR 1195.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

5Growth: AI-driven 5G for Automation in Vertical Industries

  • Papagianni, Chrysa
  • Murillo, Pablo
  • Mangues-Bafalluy, Josep
  • Bermudez, Pedro
  • Barmpounakis, Sokratis
  • Vleeschauwer, Danny De
  • Brenes, Juan
  • Zeydan, Engin
  • Casetti, Claudio
  • Guimarães, Carlos
  • García Saavedra, Andrés
  • Corujo, Daniel
  • Pepe, Teresa
This paper has been presented at 2020 European Conference on Networks and Communications (EuCNC)., Spurred by a growing demand for higher-quality mobile services in vertical industries, 5G is integrating a rich set of technologies, traditionally alien to the telco ecosystem, such as machine learning or cloud computing. Despite the initial steps taken in prior research projects in Europe and beyond, additional innovations are needed to support vertical use cases. This is the objective of the 5Growth project: automate vertical support through (i) a portal connecting verticals to 5G platforms (a.k.a. vertical slicer), a multi-domain service orchestrator and a resource management layer, (ii) closed-loop machine-learning-based Service Level Agreement (SLA) control, and (iii) end-to-end optimization. In this paper, we introduce a set of key 5Growth innovations supporting radio slicing, enhanced monitoring and analytics and integration of machine learning., This work has been partially supported by EC H2020 5GPPP 5Growth project (Grant 856709).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

An Intelligent Edge-based Digital Twin for Robotics

  • Girletti, Luigi
  • Groshev, Milan
  • Magalhaes Guimaraes, Carlos Eduardo
  • Bernardos Cano, Carlos Jesús
  • Oliva Delgado, Antonio de la
This paper has been presented at 2020 IEEE Globecom Workshop on Advanced Technology for 5G Plus., Digital Twin is one of the use cases targeted by the fourth industrial revolution (Industry 4.0), which, through the digitalization of the robotic systems, will enable enhanced automation and remote controlling capabilities. Building upon this concept, this work proposes a solution for an Edge-based Digital Twin for robotics, which leverages on the cloud-to-things continuum to offload computation and intelligence from the robots to the network. This imposes stringent requirements over the communication technologies which are fulfilled by relying on 5G. This solution is implemented in an E2E scenario combining the cloud-to-things continuum, 5G connectivity and intelligence capabilities and validated through a set of experimental evaluations. Results show not only that offloading the robot's functions to the edge is feasible when supported by the 5G connectivity, but also the benefits of introducing intelligence and automation., This work has been (partially) funded by H2020 EU/TW 5G-DIVE (Grant 859881) and H2020 5Growth (Grant 856709). It has been also funded by the Spanish State Research Agency (TRUE5G project, PID2019-108713RB-C52PID2019-108713RB-C52 / AEI / 10.13039/501100011033).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

Bayesian online learning for energy-aware resource orchestration in virtualized RANs

  • Ayala Romero, José A.
  • García-Saavedra, Andrés
  • Costa-Pérez, Xavier
  • Iosifidis, George
Proceedings of: IEEE International Conference on Computer Communications, 10-13 May 2021, Vancouver, BC, Canada., Radio Access Network Virtualization (vRAN) will spearhead the quest towards supple radio stacks that adapt to heterogeneous infrastructure: from energy-constrained platforms deploying cells-on-wheels (e.g., drones) or battery-powered cells to green edge clouds. We perform an in-depth experimental analysis of the energy consumption of virtualized Base Stations (vBSs) and render two conclusions: (i) characterizing performance and power consumption is intricate as it depends on human behavior such as network load or user mobility; and (ii) there are many control policies and some of them have non-linear and monotonic relations with power and throughput. Driven by our experimental insights, we argue that machine learning holds the key for vBS control. We formulate two problems and two algorithms: (i) BP-vRAN, which uses Bayesian online learning to balance performance and energy consumption, and (ii) SBP-vRAN, which augments our Bayesian optimization approach with safe controls that maximize performance while respecting hard power constraints. We show that our approaches are data-efficient and have provably performance, which is paramount for carrier-grade vRANs. We demonstrate the convergence and flexibility of our approach and assess its performance using an experimental prototype., This work was supported by the European Commission through Grant No. 856709 (5Growth) and Grant No. 101017109 (DAEMON); and by SFI through Grant No. SFI 17/CDA/4760.