Resources
Deliverables
D1.1 - Energy Consumption Ecosystem
This document outlines the work of T1.1, focusing on Future Energy Ecosystems. It aims to describe the environment where use cases will be implemented and their impact. This includes identifying stakeholders, their roles, energy dependencies, business relationships, and the value these use cases bring to each stakeholder and the broader ecosystem. The analysis utilises a general energy ecosystem framework within an ICT service delivery context, applying it to specific use cases from the Exigence project. This approach helps identify information exchanges that can reduce overall energy consumption, with a focus on energy measurements and incentive provisions.
D1.2 - Use Cases and Requirements
This document focuses on achieving carbon footprint transparency in digital services across their entire lifecycle. It examines various use cases within the ICT sector, including media streaming, cloud computing, and network infrastructure, to identify requirements for optimising energy efficiency and achieving carbon neutrality. Each use case provides a detailed scenario, service flows, and postconditions, highlighting existing technological solutions and identifying new requirements for full support. This analysis forms the basis for developing reliable methods to measure energy consumption and collect data, including defining necessary interfaces and procedures for data exchange.
D1.3 - Draft Functional Architecture
This document describes the "Draft Functional Architecture" developed by the project to address the growing environmental impact of the ICT sector, focusing on energy consumption and carbon footprint reduction. EXIGENCE takes an end-to-end approach, covering mobile networks (5G, 6G), cloud services, and end-user equipment. This deliverable outlines the architecture for monitoring, controlling, and optimising energy consumption and carbon impact across these domains. Key components include "EXIGENCE agents," modular elements responsible for energy monitoring and management within each domain. The document also defines how these agents interact to achieve cross-domain energy optimization and communication. This architecture serves as input for further work in other project work packages (WP2, WP3, and WP4).
D4.1 - Assessment Plan
This document outlines the initial assessment plan for the EXIGENCE project, which will be refined as the project progresses. It focuses on streamlining the evaluation process across different work packages (WPs) to avoid delays. While individual component evaluations are conducted within their respective WPs, this plan establishes validation criteria, timelines, methods, and techniques for assessing these components within the overall project context. This approach aims to decouple WP outputs, preventing delays caused by dependencies on other WPs. The document's core is an assessment and timeline table, detailing how each component (WP output) will be evaluated and mapped to specific project objectives.
D4.5 - Dissemination and Communication Plan
This document outlines the communication and dissemination strategy for the EXIGENCE project, designed to maximise its impact as per the Grant Agreement. It focuses on effectively communicating project information to stakeholders through a strategic approach that includes identifying target audiences, tailoring messages, and selecting appropriate channels. A detailed communication and dissemination plan will be developed to outline specific activities, tools, and support throughout the project's duration, ensuring alignment with overall objectives and the dynamic nature of communication. While communication is the focus here, the exploitation of project results will be addressed in a separate deliverable.
D4.7 - Innovation and Exploitation Plan
This document describes the EXIGENCE project's plan for innovation and exploitation (Task T4.3). It outlines how project results will be translated into marketable outcomes, emphasising business planning and stakeholder engagement to ensure feasibility and applicability for partners' growth. A key aspect is defining the project's Intellectual Property Rights (IPR) strategy to protect innovations. This deliverable presents the initial work, covering the project's approach to exploitation, identifying exploitable results, outlining exploitation routes, highlighting standardisation efforts, and defining the IPR strategy. This is an ongoing process throughout the project, with final results reported in D4.8. Partners will continuously update the project's results and ownership list, which will be reported on the Funding & Tenders Portal at the end of each reporting period.
D5.1 - EXIGENCE Project Handbook
The Project Handbook for EXIGENCE outlines the project's implementation approach, control processes, policies, and management strategy. It serves as a crucial document defining the necessary project management plans and their level of customisation. Acting as a central reference throughout the project's lifecycle, it ensures consistency and communication among project members and stakeholders. EXIGENCE utilises the European Commission's PM² Project Management Methodology, with this handbook based on the PM² artefact Project Handbook. The handbook is maintained and updated throughout the project.
D5.2 - Data Management Plan
This document outlines the Data Management Plan (DMP) for the EXIGENCE project, deliverable D5.2, as required by the Grant Agreement. It's part of the project management work package and focuses on how all data collected, processed, or generated will be managed throughout its lifecycle. The DMP ensures data is FAIR (Findable, Accessible, Interoperable, and Reusable) both during and after the project. It details what data will be collected, the methodologies and standards used, how data will be shared (or why not), and how it will be curated and preserved. Additionally, the DMP defines responsibilities related to data management and security.
Publications
Publisher: IEEE | 2024
Authors: M. Blöcher, N. Nedderhut, P. Chuprikov, R. Khalili, P. Eugster, L. Wang
Abstract –The emergence of network function virtualisation has enabled network function chaining as a flexible approach for building complex network services. However, the high degree of flexibility envisioned for orchestrating network function chains introduces several challenges to support dynamism in workloads and the environment necessary for their realisation. Existing works mostly consider supporting dynamism by re-adjusting provisioning of network function instances, incurring reaction times that are prohibitively high in practice. Existing solutions to dynamic packet scheduling rely on centralised schedulers and a priori knowledge of traffic characteristics, and cannot handle changes in the environment like link failures. We fill this gap by presenting FUMES, a reinforcement learning based distributed agent design for the runtime scheduling problem of assigning packets undergoing treatment by network function chains to network function instances. Our design consists of multiple distributed agents that cooperatively work on the scheduling problem. A key design choice enables agents, once trained, to be applicable for unknown chains and traffic patterns including branching, and different environments including link failures. The paper presents the system design and shows its suitability for realistic deployments. We empirically compare FUMES with state-of-the-art runtime scheduling solutions showing improved scheduling decisions at lower server capacity.
Published in: IEEE INFOCOM 2024 – IEEE Conference on Computer Communications
Publisher: IEEE | 2024
Authors: H. R. Chi, D. Corujo, R. L. Aguiar
Abstract – The edge computing market has been proliferating, which is foreseen to embrace more solution providers joining under the framework of 6G. Many research efforts exist, focusing on edge computing peer offloading among a single provider. However, there are limited discussions on multi-provider scenarios, which raises new challenge associated with providers’ synchronisation and property protection. Moreover, carbon footprint for multi-provider edge computing peer offloading will be even higher, which is also overlooked. Therefore, this paper provides a generic architecture and system model of full-decentralised on-demand edge computing peer offloading, specifically targeting multi-provider scenarios with providers’ protection and carbon reduction. The carbon-aware peer offloading algorithm is formulated as an optimisation process with Lagrangian modeling, which is physically embedded in edge computing servers, thus reaching consensus without data aggregation or property-sensitive data sharing. Simulation results reveal that the proposed algorithm achieves reduced carbon footprint, while maintaining a high quality of service.
Published in: 2024 IEEE 22nd International Conference on Industrial Informatics (INDIN)
DOI: 10.1109/INDIN58382.2024.10774500
Publisher: Association for Computing Machinery (ACM) | 2024
Authors: P. Wiesner, R. Khalili, D. Grinwald, P. Agrawal, L. Thamsen, O. Kao
Abstract – Federated Learning (FL) is an emerging machine learning technique that enables distributed model training across data silos or edge devices without data sharing. Yet, FL inevitably introduces inefficiencies compared to centralised model training, which will further increase the already high energy usage and associated carbon emissions of machine learning in the future. One idea to reduce FL’s carbon footprint is to schedule training jobs based on the availability of renewable excess energy that can occur at certain times and places in the grid. However, in the presence of such volatile and unreliable resources, existing FL schedulers cannot always ensure fast, efficient, and fair training.
Publisher: 6G IA | 2024
Authors: 6G-IA Vision and Societal Challenges Working Group
Summary – This whitepaper identifies main challenges in the area of operational Sustainability of 6G by contrasting the consensual 6G vision of the European Industry and the expected evolution of services and the mobile ecosystem with the lessons learnt from 5G, in the sense of the main energy consumers and the reasons for the latter. In trying to address these challenges, this whitepaper identifies several candidate enabling technologies, and more general approaches, for energy consumption and carbon dioxide emission reduction. It also identifies some potential current research and standardisation gaps, to be considered in the future work on the way towards more sustainable 6G. This whitepaper lays out a possible path towards more sustainable 6G operations with concrete recommendations for needed technological advances.
Publisher: Slovensko | 2024
Authors: Janez Sterle, Luka Koršič, Rudolf Sušnik
Abstract – The paper presents how network performance and energy consumption metrics, alongside key performance indicators (KPIs), are critical in designing sustainable 5G and 6G mobile systems. Article delves into strategies for optimising data throughput and reliability while minimising energy usage, ensuring a sustainable approach to network design and operation. The discussion includes innovative techniques for measuring and improving network efficiency, with a focus on maintaining high-quality service delivery.
Publisher: IEEE | 2024
Authors: Sripriya Srikant Adhatarao, Zoran Despotovic, Ljupco Jorguseski, Ewout Brandsma, Daniel Corujo, Artur Hecker
Abstract – Recent years have brought a consensus in the telecommunications industry that a strong reduction of overall energy consumption and CO2e emission of running networks is a critical goal that must be achieved. For this to be possible, a mindset change in the industry, in particular in the relevant standards setting organisations, such as 3GPP, is necessary. A whole new infrastructure for measuring energy consumption at various levels of granularity (e.g., per service, per user, etc.), measurement data collection and dissemination to relevant parties (e.g., end users, other domains, etc.), as well as optimised usage of available energy must be in place in the networks to enable this transition towards sustainable networks. This paper presents an approach on how this infrastructure can be standardised in the context of 3GPP. We analyse the current status of sustainability related efforts in 3GPP, identify the main gaps that need to be bridged, and provide concrete steps in terms of enhancements of existing technical specifications or creation of new ones to make the said infrastructure operational. We also report on concrete efforts made within the EU funded project EXIGENCE to realise this vision.
Published in: 2024 IEEE Conference on Standards for Communications and Networking (CSCN)
Publisher: IEEE | 2024
Auhtors: H. R. Chi, D. Corujo, A. Radwan, R. L. Aguiar
Abstract coming soon
Articles in Journal
Publisher: IEEE | 2024
Authors: N. Koursioumpas, L. Magoula, N. Petropouleas, A.-I. Thanopoulos, T. Panagea, N. Alonistioti, MA Gutierrez-Estevez, R. Khalili
Abstract – Progressing towards a new era of Artificial Intelligence (AI) – enabled wireless networks, concerns regarding the environmental impact of AI have been raised both in industry and academia. Federated Learning (FL) has emerged as a key privacy preserving decentralised AI technique. Despite efforts currently being made in FL, its environmental impact is still an open problem. Targeting the minimisation of the overall energy consumption of an FL process, we propose the orchestration of computational and communication resources of the involved devices to minimise the total energy required, while guaranteeing a certain performance of the model. To this end, we propose a Soft Actor Critic Deep Reinforcement Learning (DRL) solution, where a penalty function is introduced during training, penalising the strategies that violate the constraints of the environment, and contributing towards a safe RL process. A device level synchronisation method, along with a computationally cost effective FL environment are proposed, with the goal of further reducing the energy consumption and communication overhead. Evaluation results show the effectiveness and robustness of the proposed scheme compared to four state-of-the-art baseline solutions on different network environments and FL architectures, achieving a decrease of up to 94% in the total energy consumption.
Published in: IEEE Transactions on Green Communications and Networking
Videos
Dissemination Material
Powerpoint Presentation
Exigence Flyer
