EN FR
EN FR
STACK - 2025

2025Activity reportProject-Team​​​‌STACK

RNSR: 201722617P
  • Research‌ center Inria Centre at‌​‌ Rennes University
  • In partnership​​​‌ with:Ecole Nationale Supérieure​ Mines-Télécom Atlantique Bretagne Pays​‌ de la Loire, Nantes​​ Université, Orange SA
  • Team​​​‌ name: Software Stack for​ Massively Geo-Distributed Infrastructures
  • In​‌ collaboration with:Laboratoire des​​ Sciences du numérique de​​​‌ Nantes

Creation of the​ Project-Team: 2019 January 01​‌

Each year, Inria research​​ teams publish an Activity​​​‌ Report presenting their work​ and results over the​‌ reporting period. These reports​​ follow a common structure,​​​‌ with some optional sections​ depending on the specific​‌ team. They typically begin​​ by outlining the overall​​​‌ objectives and research programme,​ including the main research​‌ themes, goals, and methodological​​ approaches. They also describe​​​‌ the application domains targeted​ by the team, highlighting​‌ the scientific or societal​​ contexts in which their​​​‌ work is situated.

The​ reports then present the​‌ highlights of the year,​​ covering major scientific achievements,​​​‌ software developments, or teaching​ contributions. When relevant, they​‌ include sections on software,​​ platforms, and open data,​​​‌ detailing the tools developed​ and how they are​‌ shared. A substantial part​​ is dedicated to new​​​‌ results, where scientific contributions​ are described in detail,​‌ often with subsections specifying​​ participants and associated keywords.​​​‌

Finally, the Activity Report​ addresses funding, contracts, partnerships,​‌ and collaborations at various​​ levels, from industrial agreements​​​‌ to international cooperations. It​ also covers dissemination and​‌ teaching activities, such as​​ participation in scientific events,​​​‌ outreach, and supervision. The​ document concludes with a​‌ presentation of scientific production,​​ including major publications and​​​‌ those produced during the​ year.

Keywords

Computer Science​‌ and Digital Science

  • A1.1.8.​​ Security of architectures
  • A1.1.10.​​​‌ Reconfigurable architectures
  • A1.1.13. Virtualization​
  • A1.2.1. Dynamic reconfiguration
  • A1.2.2.​‌ Supervision
  • A1.2.4. QoS, performance​​ evaluation
  • A1.2.8. Network security​​​‌
  • A1.3.4. Peer to peer​
  • A1.3.5. Cloud
  • A1.3.6. Fog,​‌ Edge
  • A1.5.1. Systems of​​ systems
  • A1.6. Green Computing​​​‌
  • A2.1.7. Distributed programming
  • A2.1.10.​ Domain-specific languages
  • A2.5.2. Component-based​‌ Design
  • A2.6. Infrastructure software​​
  • A2.6.1. Operating systems
  • A2.6.2.​​​‌ Middleware
  • A2.6.3. Virtual machines​
  • A2.6.4. Ressource management
  • A3.1.2.​‌ Data management, quering and​​ storage
  • A3.1.3. Distributed data​​​‌
  • A3.1.8. Big data (production,​ storage, transfer)
  • A4.1. Threat​‌ analysis
  • A4.4. Security of​​ equipment and software
  • A4.9.​​​‌ Security supervision

Other Research​ Topics and Application Domains​‌

  • B2. Digital health
  • B4.​​ Energy
  • B4.5.1. Green computing​​​‌
  • B5.1. Factory of the​ future
  • B6.3. Network functions​‌
  • B6.4. Internet of things​​
  • B6.5. Information systems
  • B7.​​​‌ Transport and logistics
  • B8.​ Smart Cities and Territories​‌

1 Team members, visitors,​​ external collaborators

Research Scientists​​​‌

  • Adrien Lebre [Team​ leader, INRIA,​‌ Professor Detachement, until​​ Apr 2025, HDR​​​‌]
  • Thierry Coupaye [​Orange SA, Industrial member​‌, HDR]
  • Thomas​​ Hassan [Orange SA,​​​‌ Industrial member]
  • Daniel​ Balouek [INRIA,​‌ ISFP]
  • Sébastien Bolle​​ [Orange SA, Industrial​​​‌ member]
  • Abdelhadi Chari​ [Orange SA, Industrial​‌ member]
  • Philippe Raipin​​ Parved [Orange SA,​​​‌ Industrial member]

Faculty​ Members

  • Hélène Coullon [​‌IMT ATLANTIQUE, Associate​​ Professor, HDR]​​​‌
  • Carlos Gonzalez [IMT​ ATLANTIQUE, Associate Professor​‌]
  • Remous Koutsiamanis [​​IMT ATLANTIQUE, Associate​​​‌ Professor]
  • Thomas Ledoux​ [IMT ATLANTIQUE,​‌ Professor, HDR]​​
  • Jean-Marc Menaud [IMT​​ ATLANTIQUE, Professor,​​​‌ HDR]
  • Jacques Noyé‌ [IMT ATLANTIQUE,‌​‌ Associate Professor]
  • Kandaraj​​ Piamrat Lerebours [UNIV​​​‌ NANTES, Associate Professor‌, HDR]
  • Guillaume‌​‌ Rosinosky [IMT ATLANTIQUE​​, Associate Professor]​​​‌
  • Mario Südholt [IMT‌ ATLANTIQUE, Professor,‌​‌ HDR]

Post-Doctoral Fellows​​

  • Sidna Jeddou [IMT​​​‌ ATLANTIQUE, Post-Doctoral Fellow‌, from Sep 2025‌​‌]
  • Eloi Perdereau [​​IMT ATLANTIQUE, Post-Doctoral​​​‌ Fellow]

PhD Students‌

  • Tengfei An [INRIA‌​‌, until Sep 2025​​]
  • Severin Bradley Anzie​​​‌ [IMT ATLANTIQUE,‌ from May 2025]‌​‌
  • Simon Artus [ORANGE​​, CIFRE, from​​​‌ Mar 2025]
  • Lucien‌ Astie [IMT ATLANTIQUE‌​‌, from Oct 2025​​]
  • Hiba Awad [​​​‌SMILE, CIFRE,‌ until Feb 2025]‌​‌
  • Aymene Boucha [IMT​​ ATLANTIQUE, from Sep​​​‌ 2025]
  • Samia Boutalbi‌ [Ericsson, CIFRE‌​‌]
  • Christophe Dion [​​ORANGE LABS]
  • Celeste​​​‌ Precil Guimapi Guefack [‌IMT ATLANTIQUE]
  • Mouheb‌​‌ Jemai [CEA,​​ from Apr 2025]​​​‌
  • Houssem Jmal [UNIV‌ NANTES]
  • Mohammed Abdrrahim‌​‌ Lahmar [IMT ATLANTIQUE​​]
  • Wedwang Romial Menra​​​‌ [INRIA]
  • Martin‌ Molli [IMT ATLANTIQUE‌​‌]
  • Duc Thinh Ngo​​ [ORANGE, CIFRE​​​‌]
  • Lomig Piette [‌INRIA, from Oct‌​‌ 2025]
  • Gaëtan Plisson​​ [IMT ATLANTIQUE,​​​‌ from Oct 2025]‌
  • Olivia Proust [IMT‌​‌ Atlantique, from Sep​​ 2025]
  • Nathan Rabier​​​‌ [INRIA, from‌ Nov 2025]
  • Irina‌​‌ Samus [IMT ATLANTIQUE​​, from Oct 2025​​​‌]
  • Abdou Seck [‌IMT ATLANTIQUE, until‌​‌ Jun 2025]
  • Cherif​​ Si Mohammed [ADEME​​​‌]
  • Lylian Siffre [‌Kapela, CIFRE]‌​‌

Technical Staff

  • Alexis Bitaillou​​ [IMT ATLANTIQUE,​​​‌ Engineer]
  • Lucas Gazeau‌ [IMT ATLANTIQUE,‌​‌ Engineer, from Mar​​ 2025]
  • Baptiste Jonglez​​​‌ [INRIA, Engineer‌]
  • Sidi Mohammed Kaddour‌​‌ [INRIA, Engineer​​]
  • Anas Mokhtari [​​​‌IMT ATLANTIQUE, Engineer‌, until Jan 2025‌​‌]
  • Matthieu Rakotojaona Rainimangavelo​​ [IMT ATLANTIQUE,​​​‌ Engineer, from Feb‌ 2025 until Jun 2025‌​‌]

Interns and Apprentices​​

  • Hervé Mbailassem [IMT​​​‌ ATLANTIQUE, Intern,‌ from Mar 2025 until‌​‌ Aug 2025]
  • Sambo​​ Mohamadou Bachirou [IMT​​​‌ ATLANTIQUE, Intern,‌ from Mar 2025 until‌​‌ Aug 2025]
  • Martin​​ Plaud [INRIA,​​​‌ Apprentice, from Sep‌ 2025]
  • Gaëtan Plisson‌​‌ [IMT ATLANTIQUE,​​ Intern, from Feb​​​‌ 2025 until Aug 2025‌]
  • Nathan Rabier [‌​‌IMT ATLANTIQUE, Intern​​, from Apr 2025​​​‌ until Sep 2025]‌
  • Laid Rahmoune [INRIA‌​‌, Intern, from​​ May 2025 until Jul​​​‌ 2025]
  • Franc Abel‌ Zogning Tedongmouo [IMT‌​‌ ATLANTIQUE, Intern,​​ from Mar 2025 until​​​‌ Aug 2025]

Administrative‌ Assistant

  • Anne-Claire Binetruy [‌​‌INRIA]

Visiting Scientists​​

  • Lorenzo Carnevale [ University​​​‌ of Messina]
  • Maurice‌ Djibril Faye [The‌​‌ Cheikh Hamidou Kane Digital​​ University ]

2 Overall​​​‌ objectives

2.1 STACK in‌ a Nutshell

The STACK‌​‌ team addresses challenges related​​​‌ to the management and​ advanced usages of the​‌ Cloud to IoT continuum​​ (infrastructures on the Cloud,​​​‌ Fog, Edge, and IoT).​ More specifically, the team​‌ is interested in delivering​​ appropriate system abstractions to​​​‌ operate and use massively​ geo-distributed infrastructures, from the​‌ lowest to the highest​​ levels of abstraction (i.e.​​​‌ system to application development,​ respectively), and addressing crosscutting​‌ dimensions such as energy​​ or security. These infrastructures​​​‌ are critical for the​ emergence of new kinds​‌ of applications related to​​ the digitalization of the​​​‌ industry and the public​ sector (a.k.a. the Industrial​‌ Internet, smart cities, e-medicine,​​ etc. ).

2.2 Context​​​‌ & Objectives

Initially proposed​ to interconnect computers worldwide,​‌ the Internet has significantly​​ evolved to become in​​​‌ two decades a key​ element in almost all​‌ our activities. This (r)evolution​​ mainly relies on the​​​‌ progress that has been​ achieved in the computation​‌ and communication fields which​​ in turn has led​​​‌ to the well-known and​ widely spread Cloud Computing​‌ paradigm. Nowadays, most Internet​​ exchanges occur between endpoints​​​‌ ranging from small-scale devices,​ such as smart-phones, to​‌ large-scale facilities, i.e.,​​ cloud computing platforms, in​​​‌ charge of hosting modern​ information systems.

With the​‌ emergence of the Internet​​ of Things (IoT), stakeholders​​​‌ expect a new revolution​ that will push, once​‌ again, the limits of​​ the Internet, in particular​​​‌ by favouring the convergence​ between physical and virtual​‌ worlds into an augmented​​ world or cyber-physical world​​​‌. This convergence is​ about to be made​‌ possible thanks to the​​ development of minimalist sensors​​​‌ as well as complex​ industrial physical machines that​‌ can be connected to​​ the Internet through edge​​​‌ computing infrastructures. Edge computing​ is an extension of​‌ the cloud computing model​​ that consists in deploying​​​‌ a federation (or cooperation)​ of smaller data centers​‌ at the edge of​​ the network, thus closer​​​‌ to sensors, devices, machines,​ and end-users that produce​‌ and consume data 96​​, 94. This​​​‌ new kind of digital​ infrastructure, which covers resources​‌ from the “center” to​​ the extreme edge of​​​‌ the network, is expected​ to improve almost all​‌ aspects of daily life​​ and the decision processes​​​‌ in various domains such​ as industry, transportation, health,​‌ training and education. The​​ corresponding applications target the​​​‌ control and optimization of​ the business processes of​‌ most companies thanks to​​ the intensive use of​​​‌ ICT systems and real-time​ data collected by geo-distributed​‌ physical devices (video, sensors,​​  ...).

Among the obstacles​​​‌ to this new generation​ of Internet services is​‌ the development of a​​ convenient and powerful software​​​‌ stack, i.e. , a​ set of system mechanisms​‌ and software abstractions capable​​ of operating and exposing​​​‌ a significant number of​ diverse computational resources in​‌ a unified, efficient and​​ sustainable way.

In other​​​‌ words, this framework should​ allow operators, and devops,​‌ to manage the life-cycle​​ of both the digital​​​‌ infrastructure and the applications​ deployed on top of​‌ this infrastructure, throughout the​​ cloud to IoT continuum​​​‌. These include operations​ such as the initial​‌ configuration but also all​​ the reconfigurations that can​​ be required in response​​​‌ to particular events (maintenance‌ operations, equipment failures, application‌​‌ load variation, user mobility,​​ energy shortage, etc.).

The​​​‌ existing software stacks that‌ have been proposed to‌​‌ manage Cloud Computing platforms​​ are not appropriate for​​​‌ handling the specifics of‌ the next generation of‌​‌ digital infrastructure (in terms​​ of scale, heterogeneity, dynamicity,​​​‌ security threats, and energy‌ opportunities). For example, this‌​‌ infrastructure have to be​​ operated remotely and automatically​​​‌ as much as possible‌ as it will impossible‌​‌ to have human presence​​ on all locations. Due​​​‌ to their number, it‌ will be necessary to‌​‌ allow operations not on​​ a single site but​​​‌ on sets defined on‌ the fly as needed.‌​‌ Moreover, the management mechanisms​​ must have been designed​​​‌ to cope with intermittent‌ network access to the‌​‌ sites. That is to​​ say, offering on the​​​‌ one hand safety properties‌ and on the other‌​‌ hand autonomy in order​​ to allow each site​​​‌ to remain as operational‌ as possible in the‌​‌ event of network partitioning.​​ Finally, currently existing interfaces​​​‌ (APIs) should be extended‌ to turn location into‌​‌ a first-class citizen. In​​ particular, the locality aspects​​​‌ should be redefined from‌ the core system building‌​‌ blocks to the high-level​​ application programming interfaces.

The​​​‌ STACK activities cover the‌ full Cloud to IoT‌​‌ continuum, including recent challenges​​ related to the network​​​‌ dimension and urgent computing.‌ An enlargement of STACK‌​‌ core activities has started​​ with the arrivals of​​​‌ Ass. Prof Koutsiamanis, Ass.‌ Prof Piamrat and Dr.‌​‌ Balouek who respectively joined​​ the team in 2021,​​​‌ 2022 and 2023. In‌ 2024, the expertise of‌​‌ STACK has been further​​ strengthened with the arrival​​​‌ of Orange members, and‌ Ass. Prof Rosinosky and‌​‌ Ass. Prof Gonzalez.

2.3​​ Scientific Foundations

Through the​​​‌ ongoing integration of Orange‌ members, STACK consolidates its‌​‌ expertise in distributed systems,​​ networks, cyber-physical systems, IoT,​​​‌ device management, and software‌ programming as well as‌​‌ combining significant skills in​​ the design, practical development​​​‌ and evaluation of large-scale‌ systems. More precisely, our‌​‌ research activities mainly rely​​ on a set of​​​‌ scientific foundations detailed below.‌

  • (Distributed) Systems. The first‌​‌ scientific foundation of the​​ team is related to​​​‌ our strong expertise in‌ resource management and capacity‌​‌ planning of large-scale infrastructure​​ 76, 88,​​​‌ 80, 99,‌ 93, 92.‌​‌ This includes the design​​ and evaluation of system​​​‌ mechanisms and algorithms to‌ operate and use computation,‌​‌ network, storage, and IoT​​ resources in an efficient​​​‌ and sustainable manner. Our‌ knowledge is based on‌​‌ traditional as well as​​ distributed system fundamentals, covering​​​‌ virtualization technologies, storage, security,‌ energy, and distributed/parallel algorithms.‌​‌
  • Networks. Another set of​​ expertise in the team​​​‌ concerns network related topics.‌ This includes intelligent analysis‌​‌ and management in wireless​​ and mobile networks using​​​‌ artificial intelligence and machine‌ learning techniques, with particular‌​‌ focus on graph neural​​ networks 5, 31​​​‌, federated learning 15‌, 14 for traffic‌​‌ prediction and network security.​​ It also includes the​​​‌ optimisation of wireless low-power‌ and lossy networks (LLN),‌​‌ typically wireless Industrial IoT​​​‌ networks, through energy-aware network​ resource and communications scheduling​‌ and routing 79,​​ 77, 54,​​​‌ 75, 57.​ Additionally, using SDN enhances​‌ network security through fine-grained​​ security policies and efficient​​​‌ control plane/data plane management​ of complex routing decisions.​‌ This enables high-performance networks​​ and scalable management of​​​‌ large-scale environments, ensuring optimized​ resource usage and robust​‌ communication 70, 72​​.
  • Digital Twins, Network​​​‌ and Device Management. Based​ on initial expertise in​‌ IoT platforms and Cyber-Physical-Systems,​​ the management of connected​​​‌ devices and sensor data,​ and especially on distributed​‌ and autonomic architectures of​​ such platforms 89,​​​‌ 46, 45,​ the team has developed​‌ a broader vision of​​ network management and operation​​​‌ with a strong expertise​ in digital twins as​‌ a pivotal technology. This​​ includes graph-based modelling of​​​‌ digital twins 87,​ semantic modelling and ontology​‌ mapping 47, graph​​ storage distribution, federation and​​​‌ historisation 73, 74​ — and the application​‌ of these concepts and​​ technologies in different use​​​‌ cases in the domain​ of smart building (e.g.,​‌ localisation 57, dynamic​​ wireless IoT network resource​​​‌ allocation75), smart​ industry (e.g., support for​‌ reliable and low-latency wireless​​ Industrial IoT networks 79​​​‌, 77), logistics​ 84 around the Thing'in​‌ digital twin platform 64​​ (Thing'in the future​​​‌).
  • Autonomic and Self-Adaptive​ Systems. Considering the high​‌ (and ever increasing complexity)​​ of ICT systems, autonomic​​​‌ and self-adaptive policies have​ become the de facto​‌ standard for designing and​​ building large-scale systems. This​​​‌ second family includes, for​ example, research approaches that​‌ have been harnessed to​​ tackle system modularity, configuration​​​‌ and reconfiguration of dynamic​ and distributed systems, as​‌ well as retroaction and​​ autonomic loops. All these​​​‌ concepts enable administrators and​ developers to deal with​‌ various objectives such as​​ performance, high availability, low​​​‌ energy consumption, etc. STACK​  members have provided several​‌ relevant contributions in the​​ last couple of years​​​‌ 67, 68,​ 83, 69,​‌ 82, 53,​​ 55, 56.​​​‌
  • Software Engineering and Programming.​ Similarly, software engineering and​‌ advances in programming are​​ highly valuable to correctly​​​‌ design complex systems such​ as the software stack​‌ we target. Leveraging the​​ expertise of software programming​​​‌ of the team, STACK​  contributions leverage various techniques​‌ including component-based programming models​​  65, 50,​​​‌ 66, 81,​ 62, 63,​‌ event-driven 71, 97​​, data-driven and workflow​​​‌ models, as well as​ models for Utility Computing​‌ (Service Level Agreement, aka.​​ SLA) 95, and​​​‌ more generally, distributed and​ parallel programming models.
  • Experiment-Driven​‌ Research. Finally, the last​​ important domain of expertise​​​‌ of the future team​ consists in the evaluation​‌ of complex software stacks​​ at large scale through​​​‌ simulations and in-vivo experiments.​ This includes knowledge on​‌ experimental methodology, measuring/monitoring/tracing tools​​ 48 and more recently​​​‌ aspects related to software-defined​ experiments and reproducible research​‌ 59, 58.​​ Team members are also​​​‌ in charge of the​ animation of the LASCARE​‌ working group (LArge SCale​​ ARchitecture Experimentation and Simulation)​​ of the IOLab.

We​​​‌ aim at strengthening the‌ knowledge in these different‌​‌ areas through two kinds​​ of contributions: First through​​​‌ scientific articles as a‌ regular project team, and‌​‌ second, through concrete pieces​​ of software that can​​​‌ be transferred to major‌ opensource communities.

3 Research‌​‌ program

3.1 Overview

STACK​​ activities have been focused​​​‌ on the management and‌ programming of geo-distributed data‌​‌ centers with a work​​ program defined around four​​​‌ research topics as depicted‌ in Figure 1a. The‌​‌ first two ones are​​ related to the resource​​​‌ management mechanisms and the‌ programming support that are‌​‌ mandatory to operate and​​ use ICT geo-distributed resources​​​‌ (compute, storage, network). They‌ are transversal to the‌​‌ three software layers that​​ generally compose a software​​​‌ stack (System/Middleware/Application in Figure‌ 1b) and nurture each‌​‌ other (i.e.,​​ the resource management mechanisms​​​‌ will leverage abstractions/concepts proposed‌ by the programming support‌​‌ axis and reciprocally). The​​ third and fourth research​​​‌ topics are related to‌ the Energy and Security‌​‌ dimensions (both also crosscutting​​ the three software layers).​​​‌ Although they could have‌ been merged with the‌​‌ first two axes, we​​ identified them as independent​​​‌ research directions due to‌ their critical aspects with‌​‌ respect to the societal​​ challenges they represent.

0.44​​​‌ 0.54
(a) t (b)‌ b
Figure 1.a
Figure 1.b
Software stack layers‌​‌

The four research topics​​

Software stack layers

Figure​​​‌ 1: Stack Research‌ Topics

This scientific roadmap‌​‌ to address challenges related​​ to the management and​​​‌ programming of geo-distributed infrastructures‌ applies to the Cloud‌​‌ to IoT continuum and​​ continues to have significant​​​‌ scientific and socio-economic impact.‌ Hence, STACK organizes its‌​‌ activities around these four​​ crosscutting lines that form​​​‌ a unique approach.

Additionally,‌ our activities extend to‌​‌ the management of IoT​​ devices with the ultimate​​​‌ goal of covering the‌ entire Cloud-to-IoT continuum through‌​‌ a common software stack.​​

Our vision is to​​​‌ base this computing continuum‌ software stack on control‌​‌ loops following the MAPE-K​​ model1, which​​​‌ can be seen as‌ an infinite loop that‌​‌ monitors the infrastructure as​​ well as the state​​​‌ of the applications in‌ order to maintain in‌​‌ an autonomous manner the​​ expected objectives (in terms​​​‌ of performance, robustness, etc.).‌

Although it is largely‌​‌ adopted in Cloud orchestrators​​ such as Kubernetes, delivering​​​‌ a MAPE-K software stack‌ for the computing continuum‌​‌ faces multiple challenges. The​​ first one is related​​​‌ to the diversity of‌ resources to consider. KubeEdge‌​‌ 100, for instance,​​ proposes extensions to integrate​​​‌ servers and IoT devices‌ under the same framework.‌​‌ However, the supported operations​​ are rather limited as​​​‌ they only cover communication‌ between software components running‌​‌ on servers at the​​ edge and the connected​​​‌ devices. In other words,‌ the IoT devices are‌​‌ not considered in the​​ control loops. From our​​​‌ viewpoint, this weak integration‌ is linked to an‌​‌ incomplete understanding of the​​ needs that such a​​​‌ system must take into‌ account (in particular on‌​‌ the IoT device management​​ side). To favor the​​​‌ integration of both dimensions‌ into a common system,‌​‌ we aim to identify​​​‌ major structural properties as​ well as management operations​‌ necessary at the operator​​ and DevOps levels and​​​‌ to implement them when​ needed. A second important​‌ challenge is related to​​ the geo-distribution property (and​​​‌ so the intermittent network​ connectivity) of this type​‌ of infrastructure. This increased​​ complexity implies revising the​​​‌ way control loops are​ designed in order to​‌ handle frequent disconnections that​​ can occur at any​​​‌ time (for instance due​ to the low energy​‌ level of IoT devices).​​ Here, our approach is​​​‌ to combine autonomous loops​ with well-adapted formal methods​‌ to guarantee their verification​​ or to synthesize correct-by-design​​​‌ decisions. Additionally, we study​ their performance models to​‌ be able to automatically,​​ safely, efficiently and in​​​‌ a timely manner adapt​ Cloud-to-IoT infrastructures and their​‌ hosted applications according to​​ different objectives (performance, energy,​​​‌ security, etc.). Finally, we​ are working towards partitioning​‌ a Cloud-to-IoT infrastructure into​​ several areas and delivering​​​‌ the illusion of a​ single system through a​‌ federated approach: each area​​ is managed by an​​​‌ independent controller, and collaborations​ between areas are done​‌ through dedicated middleware. The​​ innovative aspect relies in​​​‌ the way of developing​ this middleware so that​‌ it is reliable in​​ spite of increasing scale​​​‌ and faults (collaborations will​ be triggered only on-demand​‌ without maintaining a global​​ knowledge base of the​​​‌ entire infrastructure).

Some of​ the research questions we​‌ address in the medium​​ term are:

  • How to​​​‌ specify and model the​ dynamics of Cloud-to-IoT infrastructures​‌ and the dynamics of​​ associated systems and applications​​​‌ in a generic way​ and how to leverage​‌ this dynamics for reconfiguration​​ purposes? In particular, this​​​‌ is done though studying​ how existing languages such​‌ as SysML, ThingML or​​ TOSCA may be revised​​​‌ for this purpose. In​ addition, we focus on​‌ the exploration of the​​ properties of safety, separation​​​‌ of concerns and efficiency​ when reconfiguring distributed systems​‌ with Concerto 56 extended​​ to IoT devices and​​​‌ to network resources. Finally,​ we address the dynamicity​‌ of systems described by​​ such Architecture Description Languages​​​‌ (ADL) considering a convergence​ between the Model@Runtime and​‌ Digital Twins approaches, i.e.​​ implementing models@runtime as digital​​​‌ twins.
  • How to design​ and deploy decentralized autonomic​‌ loops, from monitoring to​​ the execution of reconfiguration​​​‌ plans? An important challenge​ is related to the​‌ development of mechanisms capable​​ of rebuilding, on-demand, a​​​‌ knowledge base according to​ the functional and non-functional​‌ properties to be satisfied.​​ From our viewpoint, it​​​‌ is crucial to propose​ alternative approaches to avoid​‌ maintaining such a global​​ knowledge base through time​​​‌ and at the scale​ of a Cloud-to-IoT infrastructure.​‌ Regarding the monitoring/supervision of​​ the infrastructure, we study​​​‌ the latest results on​ complex event processing as​‌ well as machine learning​​ techniques. The former enables​​​‌ the triggering of actions​ based on predefined events​‌ while the latter allow​​ the management to evolve​​​‌ from reactive to predictive​ strategies. Regarding the analysis​‌ and planning phases we​​ study how to combine​​​‌ and leverage satisfiability solvers​ (SMT and SAT solvers),​‌ constraint programming, and distributed​​ algorithms. We are also​​ interested in the formal​​​‌ verification of reconfiguration procedures.‌
  • How to enhance legacy‌​‌ distributed software in an​​ easy and non-instrusive manner​​​‌. In particular, we‌ focus on handling the‌​‌ geo-distribution of applications, optimize​​ their processing or include​​​‌ security features using service‌ meshes and code injection‌​‌ techniques.
  • How to increase​​ the responsiveness of data​​​‌ analysis algorithms and accelerate‌ responses of AI-enabled scenarios‌​‌ across the Cloud-to-IoT continuum?​​ Another important challenge is​​​‌ related to the steering‌ of computation considering data‌​‌ and events measured by​​ the IoT infrastructure, coupled​​​‌ with historical information and‌ the Quality of Service‌​‌ (QoS) needed. We investigate​​ application formulations that allow​​​‌ developers to balance requirements‌ and costs, along with‌​‌ programming abstractions to define​​ policies that can react​​​‌ to unforeseen events and‌ constraints.
  • On the energy‌​‌ dimension, the main questions​​ are related to the​​​‌ generalization of the usage‌ of renewable energies in‌​‌ the Cloud-to-IoT continuum while​​ guaranteeing availability and reliability​​​‌ properties. We investigate,‌ in particular, whether energy‌​‌ harvesting devices could be​​ used at the extreme​​​‌ edge and how they‌ complicate the placement challenge‌​‌ that we largely studied​​ in a multi-cloud context.​​​‌ Besides, we are working‌ on extending our work‌​‌ to include green energy​​ awareness for users (e.g.,​​​‌ DevOps engineers, web application‌ end-users, etc.).
  • How can‌​‌ telemetry metrics collected by​​ P4 be integrated into​​​‌ a distributed architecture for‌ both local and global‌​‌ decision-making? Integrating telemetry metrics​​ collected by P4 into​​​‌ a distributed architecture enables‌ real-time monitoring and adaptive‌​‌ decision making. P4's programmable​​ data plane allows the​​​‌ extraction of fine-grained metrics‌ such as latency, packet‌​‌ loss, and throughput. These​​ metrics can be fed​​​‌ into a centralized or‌ hierarchical control system, enabling‌​‌ local decisions to optimize​​ immediate performance and global​​​‌ decisions for long-term resource‌ allocation and policy enforcement.‌​‌ By leveraging SDN or​​ ICN principles, these telemetry​​​‌ insights can improve traffic‌ management, enhance fault tolerance,‌​‌ and support scalable and​​ efficient operation across distributed​​​‌ environments.
  • Finally, on the‌ security side, we investigate‌​‌ the new threats resulting​​ from an externalized management​​​‌ of geo-distribution. This‌ includes, in particular, the‌​‌ identification of new possible​​ attack channels as well​​​‌ as counter measures to‌ guarantee a satisfactory level‌​‌ of security through the​​ whole continuum. Furthermore, we​​​‌ are making efforts to‌ extend our work on‌​‌ kernel security policies 49​​ in order to also​​​‌ take into account the‌ network dimension and ensure‌​‌ strong isolation from Cloud/Edge​​ servers to IoT devices.​​​‌

All the aforementioned research‌ questions are addressed through‌​‌ several application fields: telecommunications​​ operators and smart buildings​​​‌ in the first place‌ through this privileged partnership‌​‌ with the Orange colleagues​​ who are going to​​​‌ join this new team‌ but also in health,‌​‌ in particular, biomedical research​​ in order to allow​​​‌ the execution of analyses,‌ currently emerging, in large-scale‌​‌ geo-distributed environments.

4 Application​​ domains

Industrial/Tactile Internet/Cyber-Physical applications​​​‌ highlight the importance of‌ the computing continuum model.‌​‌ Hence, the use-cases of​​ STACK activities are driven​​​‌ and nurtured by these‌ application domains. Besides, it‌​‌ is noteworthy to mention​​​‌ that Telecom operators such​ as Orange have been​‌ among the first ones​​ to advocate the deployment​​​‌ of Fog/Edge infrastructure. The​ initial reason is that​‌ a geo-distributed infrastructure enables​​ them to virtualize a​​​‌ large part of their​ resources and thus reduce​‌ capital and operational costs.​​ As an example, several​​​‌ researchers have been investigating​ through the IOLab, the​‌ joint lab between Orange​​ and Inria, how 5G​​​‌ networks can be managed.​ We highlight that while​‌ our expertise does partially​​ include the network side,​​​‌ the main focus is​ rather on how we​‌ can deploy, locate and​​ reconfigure the software components​​​‌ that are mandatory to​ operate next generation of​‌ network/computing infrastructure. The main​​ challenges are related to​​​‌ the high dynamicity of​ the infrastructure, the way​‌ of defining Quality of​​ Service of applications and​​​‌ how it can be​ guaranteed. We expect our​‌ contributions will deliver advances​​ in location based services,​​​‌ optimized local content distribution​ (data-caching) and Mobile Edge​‌ Computing 2. In​​ addition to bringing resources​​​‌ close to end-users, massively​ geo-distributed infrastructures should favor​‌ the development of more​​ advanced networks as well​​​‌ as mobile services.

4.1​ Overview

Supporting industrial actors​‌ and open-source communities in​​ building an advanced software​​​‌ management stack is a​ key element to favor​‌ the advent of new​​ kinds of information systems​​​‌ as well as web​ applications. Augmented reality, telemedecine​‌ and e-health services, smart-city,​​ smart-factory, smart-transportation and remote​​​‌ security applications are under​ investigations. Although, STACK does​‌ not intend to address​​ directly the development of​​​‌ such applications, understanding their​ requirements is critical to​‌ identify how the next​​ generation of ICT infrastructure​​​‌ should evolve and what​ are the appropriate software​‌ abstractions for operators, developers​​ and end-users. STACK team​​​‌ members have been exchanging​ since 2015 with a​‌ number of industrial groups​​ (notably Orange Labs and​​​‌ Airbus), a few medical​ institutes (public and private​‌ ones) and several telecommunication​​ operators in order to​​​‌ identify both opportunities and​ challenges in each of​‌ these domains, described hereafter.​​

4.2 Industrial Internet

The​​​‌ Industrial Internet domain gathers​ applications related to the​‌ convergence between the physical​​ and the virtual world.​​​‌ This convergence has been​ made possible by the​‌ development of small, lightweight​​ and cheap sensors as​​​‌ well as complex industrial​ physical machines that can​‌ be connected to the​​ Internet. It is expected​​​‌ to improve most processes​ of daily life and​‌ decision processes in all​​ societal domains, affecting all​​​‌ corresponding actors, be they​ individuals and user groups,​‌ large companies, SMEs or​​ public institutions. The corresponding​​​‌ applications cover: the improvement​ of business processes of​‌ companies and the management​​ of institutions (e.g.​​​‌, accounting, marketing, cloud​ manufacturing, etc.); the development​‌ of large “smart” applications​​ handling large amounts of​​​‌ geo-distributed data and a​ large set of resources​‌ (video analytics, augmented reality,​​ etc.); the advent of​​​‌ future medical prevention and​ treatment techniques thanks to​‌ the intensive use of​​ ICT systems, etc. We​​​‌ expect our contributions favor​ the rise of efficient,​‌ correct and sustainable massively​​ geo-distributed infrastructure that are​​ mandatory to design and​​​‌ develop such applications.

4.3‌ Internet of Skills

The‌​‌ Internet of Skills is​​ an extension of the​​​‌ Industrial Internet to human‌ activities. It can be‌​‌ seen as the ability​​ to deliver physical experiences​​​‌ remotely (i.e.,‌ via the Tactile Internet).‌​‌ Its main supporters advocate​​ that it will revolutionize​​​‌ the way we teach,‌ learn, and interact with‌​‌ pervasive resources. As most​​ applications of the Internet​​​‌ of Skills are related‌ to real time experiences,‌​‌ latency may be even​​ more critical than for​​​‌ the Industrial Internet and‌ raise the locality of‌​‌ computations and resources as​​ a priority. In addition​​​‌ to identifying how an‌ Utility Computing infrastructure can‌​‌ cope with this requirement,​​ it is important to​​​‌ determine how the quality‌ of service of such‌​‌ applications should be defined​​ and how latency and​​​‌ bandwidth constraints can be‌ guaranteed at the infrastructure‌​‌ level.

4.4 e-Health

The​​ e-Health domain constitutes an​​​‌ important societal application domain‌ of the two previous‌​‌ areas. The STACK teams​​ is investigating distribution, security​​​‌ and privacy issues in‌ the fields of systems‌​‌ and personalized (aka. precision)​​ medicine. The overall goal​​​‌ in these fields is‌ the development of medication‌​‌ and treatment methods that​​ are tailored towards small​​​‌ groups or even individual‌ patients.

We have been‌​‌ working on different projects​​ since the beginning of​​​‌ STACK (e.g., PrivGen CominLabs).‌ In general, we are‌​‌ applying and developing corresponding​​ techniques for the medical​​​‌ domains of genomics, immunobiology‌ and transplantalogy (see Section‌​‌ 10).

The STACK​​ team continue to contribute​​​‌ to the e-Health domain‌ by harnessing advanced architectures,‌​‌ applications and infrastructure for​​ the Fog/Edge, Cloud/Edge, and​​​‌ Cloud/Edge/IoT continuum.

4.5 Network‌ Virtualization and Mobile Edge‌​‌ Services

Telecom operators have​​ been among the first​​​‌ to advocate the deployment‌ of massively geo-distributed infrastructure,‌​‌ in particular through working​​ groups such as the​​​‌ Mobile Edge Computing at‌ the European Telecommunication Standards‌​‌ Institute. The initial​​ reason is that a​​​‌ geo-distributed infrastructure enables Telecom‌ operators to virtualize a‌​‌ large part of their​​ resources and thus reduces​​​‌ capital and operational costs.‌ Through the Sylva project,‌​‌ we aim to expand​​ our collaborative efforts with​​​‌ Orange and other key‌ stakeholders in the network‌​‌ community. We focus on​​ exploring how Cloud-Native Functions​​​‌ (CNFs) and telco cloud‌ infrastructure can support innovative‌​‌ use cases over the​​ SLICES-FR experimental infrastructure.

4.6​​​‌ Urgent Computing

Urgent Computing‌ refers to a class‌​‌ of time-critical scientific applications​​ that leverage distributed data​​​‌ sources to facilitate important‌ decision-making in a timely‌​‌ manner. The overall goal​​ of Urgent Computing is​​​‌ to predict the outcome‌ of scenarios early enough‌​‌ to prevent critical situations​​ or to mitigate their​​​‌ negative consequences. Motivating use‌ cases refers to rapid‌​‌ response scenarios across the​​ Cloud-to-IoT Continuum, such as​​​‌ in natural disaster management,‌ which implies to gather‌​‌ the local state of​​ each device, transform it​​​‌ into a global knowledge‌ of the network, characterize‌​‌ the observed phenomenon according​​ to an applied model,​​​‌ and finally, trigger appropriate‌ actions. The STACK team‌​‌ investigates Urgent Computing through​​​‌ the realization of a​ fluid ecosystem where distributed​‌ computing resources and services​​ are aggregated on-demand to​​​‌ support delay-sensitive and data-driven​ workflows.

5 Social and​‌ environmental responsibility

5.1 Footprint​​ of research activities

In​​​‌ addition to the international​ travels, the environmental footprint​‌ of our research activities​​ is linked to our​​​‌ intensive use of large-scale​ testbeds such as Grid'5000​‌ (STACK members are often​​ in the top 10​​​‌ list of the largest​ consumers). Although the access​‌ to such facilities is​​ critical to move forward​​​‌ in our research roadmap,​ it is important to​‌ recognize that they have​​ a strong environmental impact​​​‌ as decribed in the​ next paragraph.

5.2 Impact​‌ of research results

The​​ environmental impact of digital​​​‌ technology is a major​ scientific and societal challenge.​‌ Even though the software​​ looks to be virtual​​​‌ objects, it is executed​ on very real hardware​‌ contributing to the carbon​​ footprint. This impact materializes​​​‌ during the manufacturing and​ destruction of hardware infrastructure​‌ (estimated at 45% of​​ digital consumption in 2018​​​‌ by the Shift Project)​ and during the software​‌ use phase via terminals,​​ networks and data centers​​​‌ (estimated at 55%). Stack​ members have been studying​‌ various approaches for several​​ years to reduce the​​​‌ energy footprint of digital​ infrastructures during the use​‌ phase. The work carried​​ out revolves around two​​​‌ main axes: (i) reducing​ the energy footprint of​‌ infrastructures and (ii) adapting​​ the software applications hosted​​​‌ by these infrastructures according​ to the energy available.​‌ More precisely, this second​​ axis investigates possible improvements​​​‌ that could be made​ by the end-users of​‌ the software themselves. At​​ scale, involving end-users in​​​‌ decision-making processes concerning energy​ consumption would lead to​‌ more frugal Cloud computing.​​

In 2025, the team​​​‌ has taken part to​ the Inria Pulse challenge​‌ (i.e., “Défis”).​​ Started in Oct 2022,​​​‌ this 4 years program​ beween Inria and Qarnot​‌ Computing, with the support​​ of ADEME aims to​​​‌ develop and promote best​ practices in terms of​‌ reducing and recycling emissions​​ of intensive computing infrastructures.​​​‌

Moreover, the team pursuits​ its activities related to​‌ the Samurai platform, a​​ project to design an​​​‌ innovative hardware infrastructure for​ the scientific study of​‌ the cross-cutting issues of​​ computing infrastructures supporting artificial​​​‌ intelligence and their energy​ autonomy (see the Samurai​‌ project Section in 10​​).

Finally, the team​​​‌ has initiated and taken​ part to several debates,​‌ panels and workshops related​​ to the environmental impact​​​‌ and the illusion of​ unlimited resources of cloud​‌ computing platforms (e.g. Infrastructure​​ workshop in Dec 2025​​​‌)

6 Highlights of​ the year

Regarding scientific​‌ contributions, the team has​​ produced major results on​​​‌ the management of large-scale​ infrastructures, in particular the​‌ team continued to study​​ the relevance of AI​​​‌ technics 2, 3​, 1

On the​‌ software side, the team​​ has pursued its efforts​​​‌ on the development of​ the EnosLib library and​‌ the resulting artifacts to​​ help researchers perform experiment​​​‌ campaigns with direct contributions​ from several research engineers​‌ of the team.

On​​ the platform side, we​​ continued our effort and​​​‌ took part in the‌ different actions around the‌​‌ SLICES and SLICES-FR, see​​ Section 7.2.

Finally, we​​​‌ would like to highlight‌ our continued efforts to‌​‌ maintain the strong visibility​​ of the team on​​​‌ the convergence of telco/cloud‌ and edge-related topics. In‌​‌ particular, the team has​​ taken part to two​​​‌ european submissions (both under‌ review, HORIZON-JU-SNS-2025-01-STREAM-C-01 and HORIZON-CL4-2025-03-DATA-08).‌​‌ The second submission is​​ a major initiative with​​​‌ a €70 million budget,‌ bringing together over 60‌​‌ partners across Europe. Being​​ sollicited to take part​​​‌ to such actions are‌ prime examples of the‌​‌ recognition of our work,​​ alongside our significant involvement​​​‌ in the French Cloud‌ PEPR initiative.

6.1 Awards‌​‌

At the 25th IFIP​​ WG 6.1 International Conference​​​‌ on Distributed Applications and‌ Interoperable Systems (DAIS 2025),‌​‌ held as part of​​ DisCoTec 2025 in Lille,​​​‌ France (June 16-20, 2025),‌ the paper "Justin: Hybrid‌​‌ CPU/Memory Elastic Scaling for​​ Distributed Stream Processing" 20​​​‌ received the best-paper award‌ recognition for both the‌​‌ DAIS and the DisCoTec​​ conference. Authored by researchers​​​‌ Donatien Schmitz, Guillaume Rosinosky,‌ and Etienne Rivière the‌​‌ paper introduces Justin, a​​ novel auto-scaling system for​​​‌ distributed stream processing engines‌ like Apache Flink.

The‌​‌ ANR SeMaFoR project lead​​ by Thomas Ledoux was​​​‌ awarded the Innovation Trophy‌ in the Digital Sector‌​‌ by the Pôle Universitaire​​ d'Innovation (PUI) of Nantes​​​‌ Université (award ceremony on‌ December 18, 2025). It‌​‌ was selected by the​​ PUI’s steering committee.

7​​​‌ Latest software developments, platforms,‌ open data

7.1 Latest‌​‌ software developments

7.1.1 ENOS​​

  • Name:
    Experimental eNvironment for​​​‌ OpenStack
  • Keywords:
    OpenStack, Experimentation,‌ Reproducibility
  • Functional Description:

    A‌​‌ typical experiment workflow using​​ Enos is the sequence​​​‌ of several phases:

    -‌ enos up : Enos‌​‌ will read the configuration​​ file, get machines from​​​‌ the resource provider and‌ will prepare the next‌​‌ phase

    - enos os​​ : Enos will deploy​​​‌ OpenStack on the machines.‌ This phase rely highly‌​‌ on Kolla deployment.

    -​​ enos init-os : Enos​​​‌ will bootstrap the OpenStack‌ installation (default quotas, security‌​‌ rules, ...)

    - enos​​ bench : Enos will​​​‌ run a list of‌ benchmarks. Enos support Rally‌​‌ and Shaker benchmarks.

    -​​ enos backup : Enos​​​‌ will backup metrics gathered,‌ logs and configuration files‌​‌ from the experiment.

  • Release​​ Contributions:
    - Install a​​​‌ fixed version of Docker‌ on nodes - Add‌​‌ support for kolla-ansible 12​​ (Openstack Wallaby), which is​​​‌ now the default -‌ Add support for Debian‌​‌ 11 base image -​​ Add support for Python​​​‌ 3.10 and 3.11 -‌ Drop support for Python‌​‌ 3.7 - Update to​​ Enoslib 8 - When​​​‌ creating a configuration template,‌ use a fixed version‌​‌ of kolla-ansible (so that​​ templates are not affected​​​‌ when future versions of‌ Enos update the default‌​‌ version of kolla-ansible)
  • URL:​​
  • Publication:
  • Contact:​​​‌
    Baptiste Jonglez
  • Participant:
    4‌ anonymous participants
  • Partner:
    Orange‌​‌ Labs

7.1.2 EnOSlib

  • Keywords:​​
    Distributed Applications, Distributed systems,​​​‌ Evaluation, Grid Computing, Cloud‌ computing, Experimentation, Reproducibility, Linux,‌​‌ Virtualization
  • Functional Description:

    EnOSlib​​ is a library to​​​‌ help you with your‌ distributed application experiments on‌​‌ bare-metal testbeds. The main​​​‌ parts of your experiment​ logic is made reusable​‌ by the following EnOSlib​​ building blocks:

    - Reusable​​​‌ infrastructure configuration: The provider​ abstraction allows you to​‌ run your experiment on​​ different environments (locally with​​​‌ Vagrant, Grid’5000, Chameleon, IoT-LAB​ and more)

    - Reusable​‌ software provisioning: In order​​ to configure your nodes,​​​‌ EnOSlib exposes different APIs​ with different level of​‌ expressivity

    - Reusable services:​​ Install common services such​​​‌ as Docker, monitoring stacks,​ network emulation

    - Reusable​‌ experiment facilities: Tasks help​​ you to iterate faster​​​‌ on your experimentation workflow​

    EnOSlib is designed for​‌ experimentation purpose: benchmark in​​ a controlled environment, academic​​​‌ validation.

  • Release Contributions:

    To​ reduce dependencies, the default​‌ pip package no longer​​ includes Jupyter support.

    Add​​​‌ support for Ansible 8,​ 9 and 10

  • URL:​‌
  • Publications:
    hal-01664515,​​ hal-01689726
  • Contact:
    Mathieu Simonin​​​‌
  • Participants:
    Mathieu Simonin, 6​ anonymous participants

7.1.3 Concerto​‌

  • Name:
    Concerto
  • Keywords:
    Reconfiguration,​​ Distributed Software, Component models,​​​‌ Dynamic software architecture
  • Functional​ Description:

    Concerto is an​‌ implementation of the formal​​ model Concerto written in​​​‌ Python. Concerto allows:

    1.​ the description of the​‌ life cycle and the​​ dependencies of software components,​​​‌

    2. the description of​ a component assembly that​‌ forms the overall life​​ cycle of a distributed​​​‌ software,

    3. the automatic​ reconfiguration of a Concerto​‌ assembly of components by​​ using a set of​​​‌ reconfiguration instructions as well​ as a formal operational​‌ semantics.

  • URL:
  • Publications:​​
  • Contact:
    Hélène Coullon​
  • Participant:
    4 anonymous participants​‌
  • Partners:
    IMT Atlantique, LS2N,​​ LIP

7.1.4 StreamBed

  • Name:​​​‌
    StreamBed capacity planning for​ steam processing
  • Keywords:
    Big​‌ data, Data stream, Performance​​ measure
  • Functional Description:
    StreamBed​​​‌ is a capacity planning​ system for stream processing.​‌ It predicts, ahead of​​ any production deployment, the​​​‌ resources that a query​ will require to process​‌ an incoming data rate​​ sustainably, and the appropriate​​​‌ configuration of these resources.​ StreamBed builds a capacity​‌ planning model by piloting​​ a series of runs​​​‌ of the target query​ in a small-scale, controlled​‌ testbed. We implement StreamBed​​ for the popular Flink​​​‌ DSP engine. Our evaluation​ with large-scale queries of​‌ the Nexmark benchmark demonstrates​​ that StreamBed can effectively​​​‌ and accurately predict capacity​ requirements for jobs spanning​‌ more than 1,000 cores​​ using a testbed of​​​‌ only 48 cores.
  • News​ of the Year:
    Publication​‌ and artefact available on​​ GitHub.
  • URL:
  • Publication:​​​‌
  • Contact:
    Guillaume Rosinosky​
  • Participant:
    an anonymous participant​‌
  • Partner:
    Université Catholique de​​ Louvain (UCL), Louvain-la-Neuve,Belgium

7.1.5​​​‌ Cheops

  • Name:
    Cheops for​ the edge
  • Keywords:
    Edge​‌ Computing, Geo-distribution, Infrastructure software​​
  • Functional Description:

    Cheops handles​​​‌ the task of synchronizing​ and replicating your application​‌ resources (in the sense​​ of REST).

    The user​​​‌ will interact with Cheops​ to perform operations on​‌ resources, and Cheops will​​ then interact with replicas​​​‌ of your application to​ make sure that all​‌ resource copies eventually converge​​ to the same state.​​​‌

    It is assumed that​ operations are always associated​‌ to a specific resource.​​ Cheops gives the possibility​​​‌ to specify the exact​ distribution of each resource​‌ manually so that operators​​ can define how they​​ want them to be​​​‌ spread.

  • URL:
  • Publications:‌
  • Contact:
    Adrien Lebre​​

7.1.6 Edge-to-cloud video processing​​​‌

  • Keywords:
    Edge Computing, Video‌ analysis
  • Functional Description:

    The‌​‌ software performs distributed video​​ processing to identify animals​​​‌ in video feeds, in‌ order to warn populations‌​‌ if a dangerous animal​​ is detected.

    The primary​​​‌ goal is to provide‌ a use-case to perform‌​‌ research on edge-to-cloud infrastructure.​​ To achieve this, the​​​‌ software is instrumented with‌ many metrics.

    The software‌​‌ is composed of three​​ components:

    - video capture:​​​‌ either from a camera,‌ or by replaying a‌​‌ pre-recorded video file that​​ can be parametrized

    -​​​‌ motion detection: detects motion‌ in video feeds. If‌​‌ a motion is detected,​​ the feed is forwarded​​​‌ to the recognizer component‌

    - object recognizer: uses‌​‌ a YOLO model to​​ determine the kind of​​​‌ animal visible in the‌ video feed

    These three‌​‌ components are designed to​​ execute in different places​​​‌ in the edge-to-cloud infrastructure,‌ so that researchers can‌​‌ explore the trade-off between​​ performance, latency, transferred data,​​​‌ and quality.

  • URL:
  • Contact:
    Baptiste Jonglez
  • Partner:‌​‌
    IMT Atlantique

7.2 New​​ platforms

7.2.1 Grid'5000

Participants:​​​‌ Remous-Aris Koutsiamanis, Baptiste‌ Jonglez, Adrien Lebre‌​‌, Jean Marc Menaud​​.

Grid'5000 is a​​​‌ large-scale and versatile testbed‌ for experiment-driven research in‌​‌ all areas of computer​​ science, with a focus​​​‌ on parallel and distributed‌ computing including Cloud, HPC‌​‌ and Big Data. It​​ provides access to a​​​‌ large amount of resources:‌ 12000 cores, 800 compute-nodes‌​‌ grouped in homogeneous clusters,​​ and featuring various technologies​​​‌ (GPU, SSD, NVMe, 10G‌ and 25G Ethernet, Infiniband,‌​‌ Omni-Path) and advanced monitoring​​ and measurement features for​​​‌ traces collection of networking‌ and power consumption, providing‌​‌ a deep understanding of​​ experiments. It is highly​​​‌ reconfigurable and controllable. STACK‌ members are strongly involved‌​‌ into the management and​​ the supervision of the​​​‌ testbed, notably through the‌ steering committee or the‌​‌ SeDuCe testbed described hereafter.​​

7.2.2 PiSeDuCe

Participants: Remous-Aris​​​‌ Koutsiamanis, Baptiste Jonglez‌, Jean Marc Menaud‌​‌.

We continue to​​ manage and extend the​​​‌ PiSeDuCe platform, a deployment‌ and reservation system for‌​‌ Edge Computing infrastructures composed​​ of multiple Raspberry Pi​​​‌ Cluster started in 2020.‌ The platform is typically‌​‌ composed of a cluster​​ of 8 Raspberry Pi,​​​‌ which costs less than‌ 900 euros and only‌​‌ needs an electrical outlet​​ and a wifi connection​​​‌ for its installation and‌ configuration. Funded by the‌​‌ CNRS through the Kabuto​​ project, and in connection​​​‌ with the SLICES-FR initiative,‌ we have extended PiSeduce‌​‌ to propose a device​​ to cloud deployment system​​​‌ (from devices on Fit‌ IoTLab to servers in‌​‌ Grid'5000). PiSeDuCe and SeDuce​​ led us to submit​​​‌ the Samurai CPER proposal.‌ Recent developments have made‌​‌ the platform more performant,​​ being able to manage​​​‌ 45 worker Raspberry Pis‌ using just one controller‌​‌ Raspbery Pi. The platform​​ has also been demonstrated​​​‌ and used in a‌ hands-on tutorial at the‌​‌ "Green IT - Numérique​​ responsable" 2024 summer school​​​‌ in Nantes.

7.2.3 SAMURAI‌

Participants: Remous-Aris Koutsiamanis,‌​‌ Baptiste Jonglez, Jean​​​‌ Marc Menaud.

The​ SAMURAI (Sustainable And autonoMoUs​‌ gReen computing for AI)​​ project is currently being​​​‌ financed as a part​ of the energy and​‌ digital transition theme. The​​ project aims at reinforcing​​​‌ an innovative hardware infrastructure​ for the scientific study​‌ of the intersecting problems​​ of computing infrastructure that​​​‌ supports artificial intelligence and​ its energy autonomy. SAMURAI​‌ is focused on extending​​ SeDuCe into energy autonomy​​​‌ by adding a smart​ and clean energy storage​‌ system. Additionally, it has​​ extended the capabilities of​​​‌ the platform by adding​ AI computing nodes (servers​‌ with GPUs) for the​​ scientific study of AI​​​‌ tools. Finally, it will​ also add new sensor​‌ nodes within the Nantes​​ connected object platform (Nantes​​​‌ site of the national​ SLICES-FR platform) to support​‌ future work on embedded​​ AI as well as​​​‌ moregenerally on the Cloud-Edge-IoT​ continuum. As the majority​‌ of the hardware for​​ the IoT, Edge and​​​‌ Cloud with GPUs has​ been procured, the project​‌ is focusing on building​​ the platform and network​​​‌ integration with the SeDuCe​ and SLICES-FR platform.

7.2.4​‌ SLICES-FR/SLICES

Participants: Remous-Aris Koutsiamanis​​, Baptiste Jonglez,​​​‌ Adrien Lebre, Jean​ Marc Menaud.

In​‌ 2025, STACK participated in​​ the SLICES-PP project (Preparatory​​​‌ Phase) as part of​ the SLICES-RI European infrastructure​‌ initiative. STACK was particularly​​ active within SLICES-FR, the​​​‌ French component of the​ infrastructure. STACK Members have​‌ been involved in the​​ definition and bootstrapping of​​​‌ the SLICES-FR infrastructure. This​ infrastructure can be seen​‌ as a merge of​​ the Grid'5000 and FIT​​​‌ testbeds with the goal​ of providing a common​‌ platform for experimental Computer​​ Science (Next Generation Internet,​​​‌ Internet of things, network​ functions, clouds, HPC, big​‌ data, etc.). Adrien Lèbre​​ and Remous-Aris Koutsiamanis are​​​‌ part of the SLICES-FR​ Board (currently provisional, pending​‌ the official creation of​​ the GIS legal structure).​​​‌ Additionally, Remous-Aris Koutsiamanis and​ Baptiste Jonglez are members​‌ of the Architects Committee​​ of SLICES-FR, where they​​​‌ contributed key design documents​ focused on networking and​‌ interoperability, aligning with the​​ overall SLICES-RI design. Finally,​​​‌ Adrien Lèbre serves as​ the French scientific representative​‌ on the SLICES Interim​​ Supervisory Board, a provisional​​​‌ body awaiting the formal​ establishment of the ERIC​‌ legal structure.

8 New​​ results

8.1 Resource Management​​​‌

Participants: Daniel Balouek,​ Hélène Coullon, Remous-Aris​‌ Koutsiamanis, Adrien Lebre​​, Duc Thinh Ngo​​​‌, Kandaraj Piamrat,​ Thomas Ledoux, Guillaume​‌ Rosinosky, Cherif Si​​ Mohammed, Mario Südholt​​​‌.

The evolution of​ the cloud computing paradigm​‌ in the last decade​​ has amplified the access​​​‌ to on-demand services (economically​ attractive, easy-to-use manner, etc.).​‌ However, the current model,​​ built upon a few​​​‌ large datacenters (DCs), is​ not suited to guarantee​‌ the needs of new​​ use cases, notably the​​​‌ boom of the Internet​ of Things (IoT). To​‌ better respond to the​​ new requirements (in terms​​​‌ of delay, traffic, etc.),​ compute and storage resources​‌ should be deployed closer​​ to the end-user, forming​​​‌ with the national and​ regional data centers a​‌ new computing continuum. The​​ question is then how​​ to manage such a​​​‌ continuum to provide end-users‌ the same services that‌​‌ made the current cloud​​ computing model so successful.​​​‌ In 2025, we have‌ continued our effort to‌​‌ answer this question and​​ delivered multiple contributions, including​​​‌ additional results related to‌ Urgent Computing (see Section‌​‌ 4).

Network resource​​ management:

In telecommunications, growing​​​‌ users and devices in‌ next-generation networks (beyond 5G)‌​‌ intensify traffic demands, stressing​​ limited resources. O-RAN architectures​​​‌ enable flexible and cost-efficient‌ networks but introduce complex,‌​‌ large-scale resource management challenges.​​ Our proposed Graph-Augmented PPO​​​‌ (GPPO)31, which‌ combines Graph Neural Networks‌​‌ for topology-aware representations with​​ action masking to jointly​​​‌ optimize functional splits and‌ virtualized unit placement. Experiments‌​‌ on small- and large-scale​​ O-RAN scenarios show that​​​‌ GPPO outperforms state-of-the-art methods,‌ achieving up to 18%‌​‌ lower deployment cost and​​ 25% higher reward with​​​‌ perfect reliability. Additionally, cellular‌ traffic prediction faces challenges‌​‌ like dynamic base station​​ deployment. Network Digital Twins​​​‌ (NDT) rely on accurate‌ traffic forecasting, but existing‌​‌ spatiotemporal models often depend​​ on fixed graphs, limiting​​​‌ adaptability in dynamic networks.‌ Our proposed Flex+ 5‌​‌, an inductive graph-based​​ model that predicts eNodeB​​​‌ traffic using local k-hop‌ spatial correlations and temporal‌​‌ features, enabling operation on​​ unseen nodes and in​​​‌ data-scarce settings. Experiments on‌ large-scale cellular data show‌​‌ 5.9% accuracy improvement in​​ inductive scenarios, 22% error​​​‌ reduction with only 3‌ days of training data,‌​‌ and up to 10×​​ faster inference via knowledge​​​‌ distillation without loss of‌ accuracy. These efforts are‌​‌ part of the work​​ that have been done​​​‌ since a decade and‌ presented during the HDR‌​‌ of Kandaraj Piamrat 38​​. Finally, on the​​​‌ resource-constrained side of networks,‌ reliable communication in Low-power‌​‌ and Lossy Networks (LLNs)​​ is a key requirement​​​‌ for Industrial IoT, yet‌ single-path RPL often fails‌​‌ to provide deterministic reliability​​ and latency guarantees under​​​‌ lossy or congested conditions.‌ While multipath RPL combined‌​‌ with PAREO functions (ARQ,​​ packet replication and elimination,​​​‌ overhearing) can improve delivery‌ by leveraging controlled redundancy,‌​‌ its efficiency crucially depends​​ on how Preferred and​​​‌ Alternative Parents are selected,‌ as overly permissive choices‌​‌ may trigger excessive flooding​​ and energy waste. In​​​‌ this context, we proposed‌ ODeSe (On-Demand Selection) 16‌​‌, a two-hop parent​​ selection algorithm that aligns​​​‌ relay choices across nodes‌ at the same DODAG‌​‌ layer by encouraging shared​​ upstream relays, while dynamically​​​‌ adjusting parent assignments when‌ necessary (and falling back‌​‌ to Soft Common Ancestor​​ selection when no suitable​​​‌ alignment is possible). Experiments‌ on a 32-node, five-hop‌​‌ topology with 50% link​​ quality show that ODeSe​​​‌ reaches 99.14% packet delivery‌ while reducing redundant transmissions,‌​‌ achieving comparable reliability to​​ the most permissive baseline​​​‌ but with lower power‌ consumption (0.09 vs. 0.10‌​‌ mW in the evaluated​​ setting).

Data resource management:​​​‌

Big data stream processing‌ approaches often do not‌​‌ consider the fact that​​ state has to be​​​‌ persisted in jobs: they‌ mainly consider CPU-bound jobs.‌​‌ In this context, following​​ our previous work on​​​‌ capacity planning estimation 91‌, where we pinpointed‌​‌ the non-linear behaviour of​​​‌ join-based jobs, and proved​ the behaviour of jobs​‌ is predictable at very​​ high rates with simple​​​‌ regression methods, we have​ done an analysis of​‌ the kind of jobs​​ that are used in​​​‌ a typical industrial data​ lake provider 19.​‌ Our findings are that​​ queries are highly heterogeneous​​​‌ in size and complexity,​ and that there is​‌ an high quantity of​​ low complexity jobs involving​​​‌ high state needs. Benchmarks​ currently used in the​‌ community do not consider​​ this fact. We have​​​‌ also proposed a new​ autoscaler 20 for Apache​‌ Flink permitting to take​​ into account CPU and​​​‌ memory needs for the​ aforementioned scenarios. It identifies​‌ memory pressures and chooses​​ whether scaling CPU or​​​‌ memory should be done.​

The second activity on​‌ data management is related​​ to the Inria/Qarnot computing​​​‌ Pulse Challenge. More precisely,​ we explored data replication,​‌ a key strategy for​​ improving system performance in​​​‌ geo-distributed computing environments. While​ replication can improve efficiency,​‌ naive strategies often result​​ in excessive and unnecessary​​​‌ data transfers, leading to​ inefficient resource utilization, particularly​‌ in infrastructures interconnected via​​ heterogeneous network links. To​​​‌ address these limitations, we​ have proposed a replication​‌ strategy that estimates the​​ utility of each potential​​​‌ replica before deployment. The​ approach has been first​‌ evaluated in homogeneous environments​​ and then extended to​​​‌ heterogeneous settings with varying​ network characteristics. Simulation results​‌ show that our method​​ significantly reduces data transfers​​​‌ while maintaining high execution​ efficiency, achieving a balanced​‌ trade-off between performance and​​ resource consumption 22,​​​‌ 32, 44.​

(Bio)medical analyses:

In the​‌ course of our long-running​​ cooperation with researchers from​​​‌ Nantes University Hospital (CHU​ Nantes), we have worked​‌ on clinical data for​​ lung transplantations in order​​​‌ to develop medical analyses​ for the survival rate​‌ after such transplantations. Although​​ not directly aligned with​​​‌ STACK’s objectives, the group’s​ expertise, in particular on​‌ data management has proven​​ valuable in finding the​​​‌ right techniques to analyse​ such a large data​‌ repository.

The main limitation​​ to long-term lung transplant​​​‌ (LT) survival is chronic​ lung allograft dysfunction (CLAD),​‌ which leads to irreversible​​ lung damage and significant​​​‌ mortality. Individual factors can​ impact CLAD, but no​‌ large genetic investigation has​​ been conducted to date.​​​‌ in 2024, we have​ established the multicentric Genetic​‌ COhort in Lung Transplantation​​ (GenCOLT) biobank from a​​​‌ rich and homogeneous sub-part​ of COLT cohort 52​‌. We continued this​​ work in 2025. Precisely,​​​‌ we applied statistical data​ analyses and learning techniques​‌ to the data bank​​ on lung transplantations. This​​​‌ work that has been​ presented as part of​‌ Simon Brocard's PhD thesis,​​ has resulted, in particular,​​​‌ in the first definition​ of links between two​‌ gene markers and graft​​ rejection complications 43.​​​‌

Urgent Computing:

Urgent computing​ scenarios describe challenges in​‌ promptly responding to changes​​ in the Edge-Cloud Continuum​​​‌ or adapting the quality​ of service (QoS) constraints​‌ of the application. This​​ line of work is​​​‌ highly motivated by decision-making​ systems or natural disaster​‌ case studies such as​​ earthquake early warning and​​ wildfires.

In 2025, we​​​‌ proposed an approach focusing‌ on leveraging Artificial Intelligence‌​‌ for Data-Driven Natural Disaster​​ Management 21, to​​​‌ facilitate dynamic resource management‌ and adaptive system modeling,‌​‌ thereby addressing the decision-making​​ challenges posed by disaster​​​‌ scenarios. We then focused‌ on anticipating system states‌​‌ in such architectures by​​ predicting CPU utilization in​​​‌ the computing continuum 9‌. We extending this‌​‌ work by implementing transfer​​ learning from Virtual Machines​​​‌ to Containers using Transformers‌ 18. This work‌​‌ enables the system to​​ anticipate CPU states with​​​‌ very little prior information,‌ and thus accelerate decision-making‌​‌ for time-critical applications.

In​​ parallel tasks, we contributed​​​‌ to a systematic mapping‌ study to understand the‌​‌ different uses of AI​​ in microservices lifecycle 4​​​‌. Outcomes of this‌ study highlight the impact‌​‌ of AI during the​​ design and deployment phases​​​‌ of microservices-based implementation, along‌ with on-going practises. Finally,‌​‌ we targeted adaptation mechanisms​​ along two dimensions: (1)​​​‌ software variability for the‌ potential of capturing an‌​‌ extensive number of configuration​​ with the ambition to​​​‌ manage advanced tradeoffs between‌ cost and quality 17‌​‌, 30, and​​ (2) feedback-oriented architecture for​​​‌ Application-level observability 23.‌

8.2 Programming Support

Participants:‌​‌ Daniel Balouek, Hélène​​ Coullon, Thomas Ledoux​​​‌, Jacques Noyé,‌ Antoine Omond, Eloi‌​‌ Perdereau, Jolan Philippe​​, Hiba Awad,​​​‌ Mario Südholt, Divi‌ De Lacour.

Fog‌​‌ Modeling:

Fog Computing moves​​ some Cloud functions closer​​​‌ to where data is‌ generated. This approach cuts‌​‌ bandwidth use, reduces delays,​​ and minimizes data transfers.​​​‌ However, designing and building‌ Fog systems is complex‌​‌ and often leads to​​ mistakes.

To address this,​​​‌ we can apply software‌ engineering best practices, such‌​‌ as verifying system properties​​ before deployment. Previous research​​​‌ has focused on checking‌ non-functional aspects of Fog‌​‌ systems at early stages.​​ In 7, we​​​‌ introduce VeriFogOps, an approach‌ that automatically selects the‌​‌ right deployment tools based​​ on Quality of Service​​​‌ (QoS) needs and then‌ creates CI/CD pipelines to‌​‌ support Fog system deployment.​​ We tested and validated​​​‌ VeriFogOps with two real-world‌ use cases, using different‌​‌ QoS solutions and deployment​​ tools. Developed with our​​​‌ industry partner Smile, this‌ work helps support the‌​‌ entire lifecycle of Fog​​ systems.

In 2025, Hiba​​​‌ Awad defended her PhD‌ thesis 33 on the‌​‌ subject of this important​​ question dealing with Quality​​​‌ of service assurance before‌ deployment of Fog systems‌​‌ with model-based engineering and​​ DevOps.

Configuration languages:

 In​​​‌ 2025, in the context‌ of the PEPR Taranis,‌​‌ we have pursued our​​ work on the anatomy​​​‌ on configuration languages, in‌ collaboration with Philippe Merle‌​‌ (Spirals), as well as​​ on the semantics of​​​‌ the configuration language CUE.‌

Software deployment/reconfiguration:

 For a‌​‌ few years, the team​​ has been working on​​​‌ deployment and dynamic reconfiguration‌ of distributed software systems‌​‌ through the Concerto tool​​ suite 56, 90​​​‌, 85, 60‌, 86. Compared‌​‌ to the literature, Concerto​​ is a component model​​​‌ closer to Infrastructure-as-Code (IaC)‌ approaches of the DevOps‌​‌ community (as Aeolus 61​​​‌). In Concerto, the​ lifecycle of pieces of​‌ software (application or infrastructure)​​ are modeled in a​​​‌ programmable manner with fine​ grain dependencies, enhancing the​‌ flexibility and speed of​​ deployments and management procedures​​​‌ (i.e., reconfiguration) compared to​ existing approaches. Notably, the​‌ PhD of Antoine Omond​​ 37 and the HDR​​​‌ of Hélène Coullon 34​ has been defended in​‌ 2025.

In 2024, while​​ not being published yet,​​​‌ we have started to​ work more directly on​‌ Infrastructure-as-Code tools and languages.​​ After working on Concerto,​​​‌ Concerto-D and Ballet (i.e.,​ well constructed research models​‌ and prototypes), we think​​ it is important to​​​‌ start from existing production​ and more complex languages,​‌ with the aim of​​ bringing together both sides.​​​‌ First, in collaboration with​ Daniel Sokolowski (who visited​‌ the team in December​​ 2023) and Guido Salvaneschi,​​​‌ and in the context​ of the PEPR Cloud​‌ (taranis project), Eloi Perdereau​​ and Hélène Coullon are​​​‌ working on a submission​ arround the formal semantics​‌ and verification of Terraform​​ and Pulumi (provisioning tools).​​​‌ Second, the Inria transfer​ action (ADT) project CoAnsible​‌ has started to develop​​ an Ansible extension that​​​‌ uses Concerto as a​ coordination backend. Finally, the​‌ ANR project For-CoaLa coordinated​​ by Hélène Coullon in​​​‌ collaboration with Frédéric Loulergue​ (University of Orléans) has​‌ started. In this context​​ the PhD of Olivia​​​‌ Proust focuses on the​ formal semantics of Ansible​‌ and CoAnsible (including Concerto)​​ to verify general theorems​​​‌ on associated languages. For​ now, none of these​‌ initiative got accepted papers​​ but we are very​​​‌ active. Jolan Philippe, who​ is a former PhD​‌ student and postdoc of​​ the team is now​​​‌ associate professor at the​ University of Orléans and​‌ has join the project.​​

In addition to this,​​​‌ in the context of​ the PEPR Cloud (Taranis​‌ project), in collaboration with​​ Christian Perez, we co-supervise​​​‌ the postdoc of Quentin​ Guilloteau. From this collaboration​‌ we also have submitted​​ a first attempt to​​​‌ add the component model​ Legato on top of​‌ Concerto. The goal of​​ Legato is to handle​​​‌ heterogeneous deployment environments and​ encapsulate them in a​‌ hierarchical way. This model​​ also leads to a​​​‌ new Nested Doll Placement​ problem introduced in the​‌ submission and solved with​​ the Gurobi interger linear​​​‌ programming solver. The CIFRE​ PhD of Simon Artus​‌ also targets to extend​​ Concerto or Legato to​​​‌ be able to handle​ heterogeneous hardware, in particular​‌ deployment on devices with​​ specific protocols.

Declarative learning:​​​‌

We have motivated the​ need for a generic​‌ definitional framework and implementation​​ support for transfer learning.​​​‌ We have then introduced​ Generic Transfer Learning (GTL).​‌ GTL supports the declarative​​ definition of transfers through​​​‌ neural network transformations and​ dataset manipulations and includes​‌ corresponding Python implementation support.​​ We have also presented​​​‌ a case study demonstrating​ how to define and​‌ implement a transfer using​​ GTL in the health​​​‌ domain. 11

In the​ domain of distributed learning,​‌ we have provided a​​ first version of a​​​‌ model for monitoring techniques​ dedicated to support distributed​‌ learning algorithms. ţhat .ite...boucha:hal-05460554​​

8.3 Energy-aware computing

Participants:​​ Remous-Aris Koutsiamanis, Thomas​​​‌ Ledoux, Jean-Marc Menaud‌, Guillaume Rosinosky,‌​‌ Thierry Coupaye, Lylian​​ Siffre.

The activities​​​‌ on this axis are‌ mainly related to the‌​‌ design, development and deployment​​ of the SAMURAI project(​​​‌7.2.3), a testbed‌ that will allow researchers‌​‌ to investigate energy related​​ challenges over the computing​​​‌ continuum (from the Cloud‌ to IoT devices/cyber physical‌​‌ systems).

In 98,​​ we propose to address​​​‌ the inherent complexities of‌ effective application scheduling in‌​‌ cloud environments by focusing​​ on reducing operational costs​​​‌ more specifically in terms‌ of energy consumption and‌​‌ application migration expenses. In​​ this intent, we propose​​​‌ a two-stage decision-making process‌ supplemented by heuristics and‌​‌ metaheuristics. Extensive experiments executed​​ on a public dataset​​​‌ and industry-standard contests demonstrate‌ the significant advancements provided‌​‌ by our approach.

Local-first​​ Software:

The growing energy​​​‌ footprint of ICT services‌ has become a critical‌​‌ environmental concern. While current​​ approaches mainly focus on​​​‌ optimizing existing architectures, we‌ advovate in exploring a‌​‌ more fundamental transformation: moving​​ from cloud-centric architectures to​​​‌ local-first software architectures, where‌ data and computing reside‌​‌ primarily on end-user devices.​​ Through a preliminary study​​​‌ 25, we examine‌ the implications of such‌​‌ an architectural shift on​​ energy consumption. We analyze​​​‌ these impacts through three‌ concrete examples, demonstrating how‌​‌ local software approaches could​​ reduce energy consumption. Our​​​‌ analysis reveals both opportunities‌ and challenges associated with‌​‌ this architectural transformation.

In​​ 24, we go​​​‌ further. Through a controlled‌ empirical study using SeekTune,‌​‌ an open-source music recognition​​ application, we compare the​​​‌ client-side and server-side energy‌ and data consumption of‌​‌ a local processing architecture​​ versus a traditional SaaS​​​‌ model for a computationally‌ intensive task. Our results‌​‌ demonstrate that, while local​​ processing consumes more energy​​​‌ on the client endpoint,‌ the combined energy cost‌​‌ of data transmission and​​ server-side computation in the​​​‌ SaaS model often results‌ in a larger overall‌​‌ energy footprint. Furthermore, the​​ local version transmits significantly​​​‌ less data over the‌ network than the SaaS‌​‌ model.

8.4 Security and​​ Privacy

Participants: Wilmer Edicson​​​‌ Garzon Alfonso, Houssem‌ Jmal, Remous-Aris Koutsiamanis‌​‌, Kandaraj Piamrat,​​ Mario Südholt.

This​​​‌ year the STACK team‌ has provided new results‌​‌ on security and privacy​​ issues in the networking​​​‌ and biomedical domains. We‌ have developed new AI-based‌​‌ methods that support new​​ means of property analysis​​​‌ and dynamic adaptation.

Systems‌ security:

 As cloud and‌​‌ edge infrastructures increasingly follow​​ the confidential computing paradigm,​​​‌ tenants can no longer‌ assume that privileged software‌​‌ such as the host​​ OS or the hypervisor​​​‌ is trustworthy. In this‌ context, vTPMs are attractive‌​‌ trust anchors for integrity​​ measurement and secure key​​​‌ storage, but many existing‌ designs still depend on‌​‌ the host environment or​​ on a shared physical​​​‌ TPM, which both enlarges‌ the trusted computing base‌​‌ and weakens isolation. To​​ address these limitations, we​​​‌ propose a new vTPM‌ architecture that executes entirely‌​‌ inside an Intel SGX​​ enclave hosted within the​​​‌ tenant VM, providing strong‌ isolation from the underlying‌​‌ infrastructure 27. Our​​​‌ design includes a trust​ establishment mechanism based on​‌ SGX remote attestation via​​ a Privacy CA, which​​​‌ binds the vTPM identity​ to a specific enclave​‌ and prevents fake vTPM​​ and cuckoo attacks. We​​​‌ also protect the vTPM​ persistent state by sealing​‌ the NVRAM within the​​ enclave and by leveraging​​​‌ SGX monotonic counters to​ detect rollback attempts, thus​‌ mitigating NVRAM tampering and​​ state rollback. Experimental results​​​‌ show that this strong​ isolation comes with minimal​‌ overhead, and in several​​ cases improves performance compared​​​‌ to a baseline vTPM​ without SGX protection.

Network​‌ security and privacy:

 AI​​ has enabled intelligent way​​​‌ of management network security​ and privacy. Particularly, Federated​‌ Learning (FL) enables distributed​​ intelligence at the network​​​‌ edge, while semi-decentralized FL​ (SDFL) improves robustness by​‌ coordinating multiple servers instead​​ of a single central​​​‌ one. To benefit of​ both, we propose TUNE-FL​‌ 15, an adaptive​​ semi-synchronous SDFL framework that​​​‌ ensures consensus under arbitrary​ topologies and efficiently handles​‌ client heterogeneity in computation​​ and data distribution. Experiments​​​‌ on intrusion detection datasets​ show that TUNE-FL outperforms​‌ representative baselines in accuracy​​ while reducing training time​​​‌ by up to 97x.​ FL also enables privacy-preserving​‌ collaborative intrusion detection system​​ (IDS) for IoT networks​​​‌ but suffers from efficiency​ issues due to client​‌ heterogeneity and non-IID data.​​ To tackle this issue,​​​‌ we propose FLAIR 14​, an adaptive semi-synchronous​‌ FL framework that uses​​ a Decision Transformer to​​​‌ intelligently select clients based​ on current and historical​‌ information, reducing communication and​​ computation overhead. Experiments on​​​‌ IDS datasets show that​ FLAIR reduces computation and​‌ communication time by up​​ to 94% and 93%,​​​‌ respectively, while maintaining comparable​ detection accuracy to baseline​‌ methods. In addition, we​​ begin investigating hierarchical supervision​​​‌ and monitoring in the​ continuum 26 in order​‌ to tackle heterogeniy issues​​ while preserving privacy and​​​‌ optimizing overhead.

Secure and​ cooperative autonomous systems:

We​‌ have proposed a unified​​ reference architecture for cooperative​​​‌ autonomous systems (CAS), called​ RACAS, to support the​‌ design of efficient and​​ modular CASes. We have​​​‌ also shown how such​ systems can be secured​‌ on-demand. In addition, we​​ have illustrated how RACAS​​​‌ can be applied to​ an intelligent transportation system​‌ including networks of autonomous​​ and connected vehicles. 10​​​‌

9 Bilateral contracts and​ grants with industry

9.1​‌ Bilateral contracts with industry​​

Kelio (formely Bodet Software)​​​‌

Participants: Thomas Ledoux.​

The ArchOps 2 Chair​‌ (for Architecture, Deployment and​​ Administration of Agile IT​​​‌ Infrastructures) is an industrial​ chair of IMT Atlantique,​‌ in partnership with Kelio​​, an SME specialized​​​‌ in solutions for time​ and attendance management. It​‌ is dedicated to all​​ IMT Atlantique students in​​​‌ the field of IT.​ It is also a​‌ channel for the transfer​​ of high-level skills: researchers,​​​‌ experts and industrials.

In​ 2025, several activities were​‌ conducted, such as a​​ provocatively named conference ”2022-2026:​​​‌ Comment l'IA "a mangé"​ le Développement Logiciel” with​‌ approximately 120 attendees, and​​ preliminary discussions on the​​​‌ topic of a joint​ doctoral thesis with the​‌ Stack team.

9.2 Bilateral​​ grants with industry

Alterway/Smile​​

Participants: Thomas Ledoux,​​​‌ Hiba Awad.

In‌ 2020, during the preparation‌​‌ of the ANR SeMaFoR​​ project, we started a​​​‌ cooperation with Alterway/Smile,‌ an SME specialized in‌​‌ Cloud and DevOps technologies.​​ This cooperation resulted in​​​‌ a joint PhD thesis‌ (called Cifre) entitled "Quality‌​‌ of Service Assurance Before​​ Deployment of Fog Systems​​​‌ with Model-Based Engineering and‌ DevOps" started in Nov.‌​‌ 2021.

In 2025, Hiba​​ Awad defended her PhD​​​‌ thesis 33 and published‌ one last article 7‌​‌.

Kapela

Participants: Thomas​​ Ledoux, Lylian Siffre​​​‌.

Lylian Siffre started‌ his PhD in Nov.‌​‌ 2024 on the subject​​ "Impacts and Uses of​​​‌ Local-First Software for Energy‌ Optimization of IT Services",‌​‌ under a co-supervision with​​ Kapela (Constellation group), an​​​‌ SME specialized in IT‌ eco-design.

In 2025, Lylian‌​‌ Siffre published two articles​​ in international conferences (​​​‌25, 24)‌ where he demonstrated the‌​‌ interest of local-first software​​ over classic SaaS software​​​‌ to reduce the energy‌ footprint of distributed architectures‌​‌ in certain use cases.​​

Orange

Participants: Simon Artus​​​‌, Paul Bori,‌ Hélène Coullon, Divi‌​‌ de Lacour, Adrien​​ Lebre, Thomas Ledoux​​​‌, Jean-Marc Menaud,‌ Duc-Thinh Ngo, Kandaraj‌​‌ Piamrat, Mario Südholt​​.

Since 2022, Orange​​​‌ Labs and the Stack‌ team have launched several‌​‌ PhD grants.

Paul Bori​​ started his PhD in​​​‌ January 2023, with the‌ subject "Container application security:‌​‌ a programmable OS-level approach​​ to monitoring network flows​​​‌ and process executions".

Duc-Thinh‌ Ngo started his PhD‌​‌ in December 2022, on​​ the subject "Dynamic graph​​​‌ learning algorithms for the‌ digital twin in edge-cloud‌​‌ continuum", under a co-supervision​​ with Orange team in​​​‌ Rennes. He defended his‌ thesis on 27 Nov.‌​‌ 2025 36.

Divi​​ de Lacour has defended​​​‌ his PhD thesis 35‌ on 16 June 2025‌​‌ on the subject "Architecture​​ et services pour la​​​‌ protection des données pour‌ systèmes coopératifs autonomes", under‌​‌ a co-supervision with the​​ Orange team in Chatillon​​​‌ (Paris south region).

Since‌ April 2024, the STACK‌​‌ research group has officially​​ been established as a​​​‌ joint team between Inria,‌ IMT Atlantique, Nantes University,‌​‌ and Orange Labs. This​​ collaboration includes several PhD​​​‌ grants funded by the‌ three institutions. For clarity,‌​‌ we no longer distinguish​​ between PhD candidates based​​​‌ on their funding source‌ (Inria, IMT Atlantique or‌​‌ Orange) in this context.​​

Ericsson

Participants: Samia Boutalbi​​​‌, Mario Südholt,‌ Remous-Aris Koutsiamanis.

Samia‌​‌ Boutalbi started her PhD​​ in January 2022, on​​​‌ the subject "Secure deployment‌ of micro-services in a‌​‌ shared Cloud RAN/MEC environment",​​ under a co-supervision with​​​‌ the Ericsson team in‌ Paris. She has finalized‌​‌ her PhD manuscript 51​​ this year and her​​​‌ defense is planned for‌ March/April 2026.

10 Partnerships‌​‌ and cooperations

10.1 International​​ initiatives

10.1.1 Inria associate​​​‌ team not involved in‌ an IIL or an‌​‌ international program

RANMA -​​ Resource Adaptation and Network​​​‌ Management for continuum computing‌ Applications with the National‌​‌ Institute of Informatics (Tokyo,​​ Japan) from 2025 to​​​‌ 2027. The objective of‌ the RANMA associate team‌​‌ will focus on contributing​​​‌ to the programming support​ and the resource management​‌ of infrastructure and services​​ across the network and​​​‌ the application layers of​ the Computing Continuum. Led​‌ by D. Balouek and​​ K. Piamrat.

10.2 International​​​‌ research visitors

10.2.1 Visits​ of international scientists

Other​‌ international visits to the​​ team
Maurice Djibril Faye​​​‌
  • Status
    (Assistant professor)
  • Institution​ of origin: Université numérique​‌ Cheikh Hamidou KANE
    item[Country:Senegal]​​
  • Dates:
    Dec 1st to​​​‌ Dec 21st
  • Context of​ the visit:
    Natural disasters,​‌ conflicts, and public emergencies​​ in Africa often disrupt​​​‌ communication networks, delaying responses,​ and putting lives at​‌ risk. Emergency data is​​ also vulnerable to tampering​​​‌ and interception. The goal​ of this visit is​‌ to explore research themes,​​ and possibilities of collaboration​​​‌ for rapid, resilient, and​ trustworthy emergency communications.
  • Mobility​‌ program/type of mobility:
    Research​​ stay
Theodoros TSIOLAKIS
  • Status:​​​‌
    (PhD candidate)
  • Institution of​ origin:
    Democritus University of​‌ Thrace item[Country:] Greece
  • Dates:​​
    12/05/2025 - 21/05/2025
  • Context​​​‌ of the visit:
    Energy​ efficiency of machine learning​‌ implementations within geo-distributed systems.​​
  • Mobility program/type of mobility:​​​‌
    ERASMUS+ Short Term Mobility​

10.3 European initiatives

10.3.1​‌ Horizon Europe

SLICES-PP /​​ SLICES-RI (Scientific Large Scale​​​‌ Infrastructure for Computing/Communication Experimental​ Studies - Research Instrument)​‌

Participants: Adrien Lebre,​​ Remous-Aris Koutsiamanis.

The​​​‌ STACK Research team is​ actively involved in SLICES-RI,​‌ the European initiative aimed​​ at developing a large-scale​​​‌ research infrastructure for digital​ sciences. This project brings​‌ together partners from 16​​ countries, including France, Greece,​​​‌ Poland, Norway, the Netherlands,​ Italy, and Switzerland, all​‌ committed to contributing resources​​ and expertise.

Members of​​​‌ our team are currently​ actively participating in the​‌ Preparatory Phase (SLICES-PP) of​​ the project, helping to​​​‌ shape its direction and​ implementation. Our goal is​‌ to continue this involvement​​ at both the European​​​‌ level and the national​ level via SLICES-FR, contributing​‌ to the creation of​​ a flexible platform that​​​‌ supports large-scale experimental research​ in areas like networking​‌ protocols, radio technologies, and​​ cloud and edge computing​​​‌ architectures.

Once established, SLICES-RI​ aims provide researchers across​‌ Europe with long-term access​​ to advanced computing, storage,​​​‌ and network resources. This​ infrastructure will facilitate experiments​‌ in various domains, including​​ information theory, networking, distributed​​​‌ systems, and software engineering,​ thereby enabling experimentation and​‌ innovation in multiple domains​​ such as smart cities,​​​‌ e-health, industrial internet, transport,​ and energy solutions.

SEED:​‌ Training the next generation​​ of technological scientists to​​​‌ achieve Societal, Energy, Environmental,​ industrial and Digital transitions​‌

Participants: Mario Südholt [Coordinator]​​.

SEED, which stands​​​‌ for Societal, Energy, Environmental,​ industrial and Digital transitions,​‌ is a 60-month interdisciplinary,​​ international and intersectoral doctoral​​​‌ training programme offered by​ IMT Atlantique and co-funded​‌ by the European Union.​​ Its overall budget, managed​​​‌ entirely by IMT Atlantique,​ amounts to 8 M€.​‌ The programme itself is​​ designed to nurture four​​​‌ key dimensions: thesis interdisciplinarity,​ internationality, cross-sector experience, and​‌ promotion of innovation. It​​ offers 40 fully funded​​​‌ early-stage researcher (ESR) positions​ within three different tracks.​‌ Each track builds on​​ the same fundamental excellence​​​‌ trainings implementing a 4i​ approach (Interdisciplinarity, Internationality, Intersectorial,​‌ Innovation), while providing a​​ different degree of mobility​​ and focus. Further information​​​‌ at: SEED website.‌

10.3.2 Other european programs/initiatives‌​‌

DI4SPDS (Distributed Intelligence for​​ Enhancing Security and Privacy​​​‌ of Decentralised and Distributed‌ Systems)

Participants: Houssem Jmal‌​‌, Kandaraj Piamrat.​​

Decentralised systems face challenges​​​‌ from sophisticated cyber-attacks that‌ evolve and propagate to‌​‌ disrupt different parts. Additionally,​​ communication overhead makes implementing​​​‌ authentication and access control‌ complex. Existing approaches unlikely‌​‌ provide effective access control​​ and multi-stage attack detection​​​‌ due to limited event‌ capture and information correlation.‌​‌ This project offers a​​ framework to improve security​​​‌ and privacy of decentralised‌ systems through cross-domain access‌​‌ control, collaborative intrusion detection,​​ and dynamic risk management​​​‌ considering resource consumption. It‌ facilitates subsystem collaboration to‌​‌ prevent widespread disruption from​​ attacks and share threat​​​‌ awareness. The project will‌ develop methods and prototypes‌​‌ utilizing blockchain, federated learning,​​ and multi-agent architecture to​​​‌ enhance access control, detection,‌ risk management, and response‌​‌ capabilities. The consortium is​​ composed of four partners:​​​‌ Nantes University, LUT University‌ (Finland), Universidad de Castilla‌​‌ - La Mancha (Spain),​​ and Firat University (Turkey).​​​‌ LUT is the coordinator‌ of the project DI4SPDS‌​‌ has been accepted in​​ July 2023, started from​​​‌ March 2024 for 36‌ months, with an allocated‌​‌ budget of 874k€ (230K€​​ for Stack).

DISCOVER-US: Collaboration​​​‌ with NSF on fundamental‌ research on new concepts‌​‌ for distributed computing and​​ swarm intelligence

Participants: Daniel​​​‌ Balouek.

DISCOVER-US represents‌ a collaboration between EU‌​‌ and US research institutions,​​ focusing on advancing distributed​​​‌ computing and swarm intelligence.‌ At the core of‌​‌ DISCOVER-US is the commitment​​ to develop a practical​​​‌ infrastructure that supports a‌ dynamic research ecosystem. The‌​‌ stack team has significantly​​ contributed to a vision​​​‌ white paper highlighting requirements‌ and potential approaches for‌​‌ next-generation distributed computing. This​​ work is aligned with​​​‌ key objectives from Horizon‌ Europe and the US‌​‌ National Science Foundation, driving​​ progress in cloud-to-edge processing​​​‌ technologies, AI, and cybersecurity.‌

10.4 National initiatives

10.4.1‌​‌ ANR

NET4AI (Network Accelleration​​ for Generative Artificial Intelligence)​​​‌

Participants: Remous-Aris Koutsiamanis [coordinator]‌, Mario Südholt.‌​‌

Generative AI now relies​​ on very large foundation​​​‌ models whose training and‌ inference stretch compute clusters‌​‌ and—crucially—the interconnect fabric. At​​ scale, collective communications, congestion,​​​‌ and failures/stragglers limit Model‌ FLOPs Utilization (MFU) far‌​‌ more than raw FLOPs.​​ NET4AI addresses this by​​​‌ making the AI fabric‌ (network + compute) explicitly‌​‌ observable and controllable for​​ GenAI workloads, so that​​​‌ orchestration and communications adapt‌ to load, faults, and‌​‌ topology rather than treating​​ the network as a​​​‌ black box. The project‌ pursues three objectives: (1)‌​‌ High-accuracy AI fabric awareness​​ via advanced monitoring and​​​‌ workload–infrastructure co-observation to feed‌ scheduling and resilience decisions;‌​‌ (2) Optimized traffic scheduling​​ algorithms: design of​​​‌ efficient collectives tightly integrated‌ with adaptive routing, traffic‌​‌ control, fast failover, packet​​ scheduling, and network/compute-aware orchestration​​​‌ to improve completion time,‌ stability, reliability and fairness;‌​‌ (3) Large-scale evaluation and​​ experimentation with new metrics​​​‌ and tools that assess‌ end-to-end network and power‌​‌ efficiency, combining scalable simulation​​ and cutting-edge testbeds (SLICES-RI).​​​‌ The consortium is composed‌ of five partners: Laboratoire‌​‌ CEDRIC (CNAM), Laboratoire Informatique​​​‌ d’Avignon Université (AU), LS2N​ - IMT Atlantique (STACK​‌ team), Scalnyx, and Huawei.​​ NET4AI started in April​​​‌ 2025 for a period​ of 48 months, is​‌ supported by ANR under​​ AAPG 2024 (PRCE), and​​​‌ has an overall budget​ of 934K€ (203K€ for​‌ STACK). See the NET4AI​​ web site for more​​​‌ information.

For-CoaLa (Formalization of​ Configuration Languages)

Participants: Hélène​‌ Coullon [coordinator], Olivia​​ Proust.

Large distributed​​​‌ software systems (applications or​ infrastructures) are now ubiquitous,​‌ with component-based systems (e.g.,​​ service-oriented architectures or microservices)​​​‌ offering a convenient way​ to structure large systems,​‌ in particular distributed systems​​ deployed in the Cloud,​​​‌ in the core, or​ at the edge of​‌ the network. DevOps operations,​​ that include system configurations​​​‌ and reconfigurations, are required​ to handle various kinds​‌ of scenarios such as​​ fault tolerance, scalability, software​​​‌ updates, or various optimizations,​ etc. Such changes may​‌ lead to faults. A​​ study of 597 unplanned​​​‌ outages that affected popular​ cloud services between 2009​‌ and 2015 found that​​ 16% of them were​​​‌ caused by a system​ upgrade.

On the one​‌ hand, many configuration tools​​ and languages exist in​​​‌ the DevOps community, some​ of them being specific​‌ to the provisioning of​​ resources in Cloud providers,​​​‌ packaging problems, containerized deployments,​ configuration of applications or​‌ infrastructures, etc. The main​​ advantage of these tools​​​‌ is their full integration​ and adoption in the​‌ DevOps community. Their disadvantage​​ is they lack both​​​‌ formal and textual specifications.​ Moreover, their contours are​‌ blurred. On the other​​ hand, many initiatives have​​​‌ been studied in academia​ to contribute to the​‌ deployment, configuration, and reconfiguration​​ of distributed software, bringing​​​‌ improvements such as expressivity,​ speed, safety, etc. Many​‌ come with precise and​​ sometimes formal definitions. However,​​​‌ they lack the breadth​ of the mainstream DevOps​‌ tools.

The goal of​​ For-CoaLa is twofold: (1)​​​‌ understand and bridge the​ gap between a popular​‌ tool from the DevOps​​ community (Ansible) and a​​​‌ tool from academia (Concerto);​ (2) improve the understanding​‌ of these languages based​​ on mechanized formal semantics​​​‌ and develop verified semantic-preserving​ cross-language transformations. See the​‌ For-CoaLa web site for​​ more information.

SeMaFoR (Self-Management​​​‌ of Fog Resources)

Participants:​ Thomas Ledoux [coordinator],​‌ Hélène Coullon, Matthieu​​ Rakotojaona Rainimangavelo.

Fog​​​‌ Computing is a paradigm​ designed to decentralize cloud​‌ infrastructure by extending computing​​ and storage resources to​​​‌ the network's edge, enabling​ their geographic distribution along​‌ with associated services. The​​ SeMaFoR project aims to​​​‌ model, design, and implement​ a generic, decentralized solution​‌ for the self-management of​​ Fog resources. The consortium​​​‌ comprises three partners: LS2N-IMT​ Atlantique (Stack, NaoMod, TASC),​‌ LIP6-Sorbonne University (Delys), and​​ Alterway/Smile (SME). The Stack​​​‌ team oversees the project​ with an allocated budget​‌ of €506,000, of which​​ €230,000 is for Stack.​​​‌

SeMaFoR, which began in​ March 2021, concluded in​‌ August 2025. The highlight​​ of the year was​​​‌ the SeMaFoR project winning​ the 2025 Innovation Trophy​‌ for the digital sector​​ at the PUI University​​​‌ of Nantes.

For more​ information, visit the SeMaFoR​‌ website: Semafor.

PicNic​​ (Transfert de grands volumes​​ de données entre datacenters)​​​‌

Participants: Jean-Marc Menaud [STACK‌ representative], Remous-Aris Koutsiamanis‌​‌, Adrien Lebre,​​ Abdou Seck, Guillaume​​​‌ Rosinosky.

Large dataset‌ transfer from one datacenter‌​‌ to another is still​​ an open issue. Currently,​​​‌ the most efficient solution‌ is the exchange of‌​‌ a hard drive with​​ an express carrier, as​​​‌ proposed by Amazon with‌ its SnowBall offer. Recent‌​‌ evolutions regarding datacenter interconnects​​ announce bandwidths from 100​​​‌ to 400 Gb/s. The‌ contention point is not‌​‌ the network anymore, but​​ the applications which centralize​​​‌ data transfers and do‌ not exploit parallelism capacities‌​‌ from datacenters which include​​ many servers (and especially​​​‌ many network interfaces –‌ NIC). The PicNic project‌​‌ addresses this issue by​​ allowing applications to exploit​​​‌ network cards available in‌ a datacenter, remotely, in‌​‌ order to optimize transfers​​ (hence the acronym PicNic).​​​‌ The objective is to‌ design a set of‌​‌ system services for massive​​ data transfer between datacenters,​​​‌ exploiting distribution and parallelisation‌ of networks flows.

The‌​‌ consortium is composed of​​ several partners: Laboratoire d'Informatique​​​‌ du Parallélisme, Institut de‌ Cancérologie de l’Ouest /‌​‌ Informatique, Institut de Recherche​​ en Informatique de Toulouse,​​​‌ Laboratoire des Sciences du‌ Numérique de Nantes, Laboratoire‌​‌ d'Informatique de Grenoble, and​​ Nutanix France.

PiCNiC will​​​‌ be running for 42‌ months (starting in Sept‌​‌ 2021 with an allocated​​ budget of 495k€, 170k€​​​‌ for STACK).

Taranis

Participants:‌ Daniel Balouek, Helene‌​‌ Coullon [STACK Representative],​​ Adrien Lebre, Thomas​​​‌ Ledoux, Jacques Noyé‌, Eloi Perdereau,‌​‌ Nathan Rabier, Guillaume​​ Rosinosky, Gaëtan Plisson​​​‌.

New infrastructures, such‌ as Edge Computing or‌​‌ the Cloud-Edge-IoT computing continuum,​​ make cloud issues even​​​‌ more complex, as they‌ add new challenges linked‌​‌ to the diversity and​​ heterogeneity of resources (from​​​‌ small sensors to data‌ centers/HPCs, from low-power networks‌​‌ to core networks), geographical​​ distribution, as well as​​​‌ increased requirements for dynamicity‌ and security, all under‌​‌ constraints such as energy​​ consumption.

To exploit these​​​‌ new infrastructures efficiently, the‌ Taranis project is based‌​‌ on a strategy aimed​​ at abstracting the description​​​‌ of the structure of‌ applications and resources in‌​‌ order to automate their​​ management even further. In​​​‌ this way, it will‌ be possible to globally‌​‌ optimize the resources used​​ with regard to multi-criteria​​​‌ objectives (price, deadline, performance,‌ energy, etc.) on both‌​‌ the user side (applications)​​ and the resource provider​​​‌ side (infrastructures). Taranis also‌ addresses the challenges of‌​‌ abstracting application reconfiguration and​​ dynamically adapting resource usage.​​​‌

The consortium is composed‌ of 6 partners (Inria,‌​‌ CNRS, IMT, University of​​ Grenoble Alpes, CEA and​​​‌ Universtity of Rennes) for‌ a budget of 7.2M€‌​‌ ( 470K€ for STACK)​​ overall.

Spirec

Participants: Mario​​​‌ Südholt [coordinator], Remous-Aris‌ Koutsiamanis [STACK Representative],‌​‌ Kandaraj Piamrat, Carlos​​ Gonzalez, Jean-Marc Menaud​​​‌.

The SPIREC project‌ will meet the challenges‌​‌ of supervising services of​​ the Cloud-Edge-IoT continuum, detecting​​​‌ their execution anomalies and‌ predicting their resource usage.‌​‌ The project aims to​​ define methods and techniques,​​​‌ notably using distributed machine‌ learning, to enable its‌​‌ efficient management, provide means​​​‌ to secure them and,​ more generally, ensure a​‌ variety of quality of​​ service properties. The partners​​​‌ will also develop software​ components and tools in​‌ order to integrate these​​ functionalities in existing infrastructures​​​‌ and applications, in particular​ SLICES, industrial systems and​‌ future software ecosystems.

The​​ consortium is composed of​​​‌ 6 partners (Inria, CNRS,​ CEA, IMT Atlantique, Télécom​‌ SudParis and Université de​​ Versailles, Saint Quentin) for​​​‌ a budget of  2,7M€​ ( 580K€ for STACK)​‌ overall.

Steel

Participants: Daniel​​ Balouek, Carlos Gonzalez​​​‌ [STACK Representative], Jeddou​ Sidna, Guillaume Rosinosky​‌, Adrien Lebre.​​

The STEEL project aims​​​‌ to provide solutions for​ efficient and secure data​‌ storage and processing on​​ cloud-based infrastructures. The consortium​​​‌ is composed of 8​ partners (Inria, CNRS, University​‌ of Grenoble, University of​​ Bordeaux, University of Rennes,​​​‌ IMT Atlantique, IMT TeraLab​ and IN2P3) for a​‌ budget of 3 millions​​ € (300K€ for STACK)​​​‌ overall. activies are organized​ around three technical work​‌ packages. A fourth work​​ package is dedicated to​​​‌ management, communication and dissemination​ of results. STACK members​‌ are involved in the​​ second wp, addressing the​​​‌ challenges related to the​ management of data sets​‌ in presence of node​​ failures and network partitions.​​​‌

CareCloud

Participants: Jean-Marc Menaud​ [STACK representative], Remous-Aris​‌ Koutsiamanis, Thomas Ledoux​​.

At a time​​​‌ when climate change is​ a growing concern, with​‌ serious consequences for people​​ and the planet worldwide,​​​‌ all sectors (transport, construction,​ agriculture, industry, etc.) must​‌ contribute to the effort​​ to reduce GHG emissions.​​​‌ Clouds, despite their ability​ to optimize processes in​‌ other sectors, are no​​ exception to this observation:​​​‌ the increasing slope of​ their GHG emissions must​‌ be reversed, or their​​ potential benefits in other​​​‌ sectors will be wiped​ out. The CARECloud project​‌ aims to drastically reduce​​ the environmental impact of​​​‌ cloud infrastructures. The consortium​ is composed of 4​‌ partners (CNRS, IMT Atlantique,​​ Inria and Univeristy of​​​‌ Paul Sabatier - Toulouse),​ for a budget of​‌ 5.5M€ ( 600k€ for​​ STACK) overall.

Jérémy Woirhaye​​​‌ began his PhD in​ October 2025, under a​‌ joint supervision between the​​ Spirals team (Romain Rouvoy)​​​‌ and Thomas Ledoux, on​ the topic "Towards an​‌ intelligent slicing of microservices​​ applications in the cloud".​​​‌

SILECS

Participants: Baptiste Jonglez​, Remous-Aris Koutsiamanis [STACK​‌ Representative], Adrien Lebre​​, Jean-Marc Menaud.​​​‌

Digital transformation relies on​ a sophisticated infrastructure of​‌ networks, computing and services.​​ The availability, reliability, performance,​​​‌ interoperability and energy efficiency​ of these systems are​‌ major challenges that the​​ digital sciences must meet​​​‌ to foster innovation, sovereignty​ and industrial competitiveness.

SILECS,​‌ the Cloud/Fog/Edge/IoT part of​​ the SLICES-FR platform, enables​​​‌ prototyping and reproducible experiments​ at all levels of​‌ the Cloud IoT continuum.​​ It meets the experimental​​​‌ needs of researchers in​ networks, systems, telecoms, IoT​‌ and other fields. The​​ main objective of SILECS​​​‌ and SLICES-FR is to​ build a tool for​‌ experimentation that fosters the​​ design of new services​​​‌ and applications in distributed​ computing, edge computing, reprogrammable​‌ wired or wireless networks​​ and IoT, using a​​ diversity of technologies on​​​‌ all aspects of the‌ data chain, software or‌​‌ hardware, to meet the​​ needs of the community.​​​‌

The consortium is composed‌ of 3 partners (Inria,‌​‌ CNRS, and IMT) for​​ a budget of 12​​​‌ millions € ( 272k€‌ for STACK) overall .‌​‌

10.4.2 PIA 4

OTPaaS​​

Participants: Farid Arfi,​​​‌ Daniel Balouek, Hélène‌ Coullon, Marie Delavergne‌​‌, Tayeb Diab,​​ Mohamed Graiet, Houssem​​​‌ Jmal, Sidi Mohammed‌ Kaddour, Remous-Aris Koutsiamanis‌​‌, Adrien Lebre [STACK​​ representative, until Feb 2024]​​​‌, Thomas Ledoux,‌ Jean-Marc Menaud, Anas‌​‌ Mokhtari, Jacques Noyé​​ [STACK representative, from Feb​​​‌ 2024], Kandaraj Piamrat‌, Eloi Perdereau,‌​‌ Matthieu Rakotojaona Rainimangavelo,​​ Mario Südholt.

The​​​‌ OTPaaS project targeted the‌ design and development of‌​‌ a complete software stack​​ to administrate and use​​​‌ edge infrastructures for the‌ industry sector. The consortium‌​‌ brought together national and​​ user technology suppliers from​​​‌ major groups (Atos /‌ Bull, Schneider Electric, Valeo)‌​‌ and SMEs / ETIs​​ (Agileo Automation, Mydatamodels, Dupliprint,​​​‌ Solem, Tridimeo, Prosyst, Soben),‌ with a strong support‌​‌ from major French research​​ institutes (CEA, Inria, IMT,​​​‌ CAPTRONIC). The project started‌ in October 2021 for‌​‌ a period of 36​​ months with an overall​​​‌ budget of 56M€ (1.2M€‌ for STACK) and was‌​‌ extended until April 2025​​ (with very limited involvement​​​‌ of STACK in 2025).‌

The OTPaaS platform objectives‌​‌ were:

  • To be built​​ on national and sovereign​​​‌ technologies for the edge‌ cloud.
  • To be validated‌​‌ by industrial demonstrators of​​ multisectoral use cases.
  • To​​​‌ be followed and supported‌ by ambitious industrialization programs.‌​‌
  • To be accompanied by​​ a massive campaign to​​​‌ promote its use by‌ SMEs / midcaps.
  • To‌​‌ integrate solutions for controlling​​ energy consumption.
  • To be​​​‌ compliant with the Gaia-X‌ ecosystem.

10.4.3 CPER

SAMURAÏ‌​‌

Participants: Jean-Marc Menaud [coordinator]​​, Remous-Aris Koutsiamanis.​​​‌

The SAMURAI (Sustainable And‌ autonoMoUs gReen computing for‌​‌ AI) infrastructure aims to​​ design an innovative hardware​​​‌ infrastructure for the scientific‌ study of the cross-cutting‌​‌ issues of computing infrastructures​​ supporting artificial intelligence and​​​‌ their energy autonomies.

This‌ project paves the way‌​‌ toward a larger infrastructure​​ at the natonial level​​​‌ in the context of‌ the SLICES-FR initiative.

The‌​‌ project started in 2022​​ for a period of​​​‌ 5 years with an‌ overall budget of 730K€‌​‌ (500K€ for STACK).

10.4.4​​ Local and regional projects​​​‌

SysMics network

Participants: Mario‌ Südholt.

SysMics is‌​‌ an integrated cluster of​​ research that is part​​​‌ of the Nantes Excellence‌ Initiative in Medecine and‌​‌ Engineering. Its main objective​​ is the development of​​​‌ new methods for precision‌ medecine, in particular, based‌​‌ on genomic analyses. In​​ this context, we have​​​‌ worked on new large-scale‌ distributed biomedical analyses and‌​‌ provided several results on​​ how to distributed popular​​​‌ statistical analyses, such as‌ FAMD-based and EM-based analyses.‌​‌

10.4.5 Inria Challenges

FrugalCloud​​ (Inria-OVHCloud)

Participants: Hélène Coullon​​​‌, Thomas Ledoux.‌

A joint collaboration between‌​‌ Inria and OVHcloud company​​ on the challenge of​​​‌ frugal cloud has been‌ launched in October 2021‌​‌ with a budget of​​​‌ 2 M€. It addresses​ several scientific challenges on​‌ the eco-design of cloud​​ frameworks and services for​​​‌ large scale energy and​ environmental impact reduction, across​‌ three axes: i) Software​​ eco-design of services and​​​‌ applications; ii) Efficiency leverages;​ iii) Reducing the impact​‌ and supporting users of​​ the Cloud.

The main​​​‌ activities of the Stack​ team were at the​‌ start of the project​​ between 2021 and 2024​​​‌ (e.g., PhD thesis of​ Pierre Jacquet).

CUPSELI (Inria-Hivenet)​‌

Participants: Guillaume Rosinosky,​​ Mario Südholt, Lomig​​​‌ Piette.

The défi​ CUPSELI is a joint​‌ collaboration between INRIA and​​ the Hivenet company. Its​​​‌ aim is to push​ the limits of distributed​‌ AI computing. Its goal​​ is to demonstrate that​​​‌ even the most demanding​ AI and Big Data​‌ applications can run efficiently​​ on heterogeneous, distributed, and​​​‌ volatile resources — while​ maintaining accuracy, ensuring privacy,​‌ and reducing environmental impact.​​ The project is structured​​​‌ amongst three axes: frugality,​ security and confidentiality, and​‌ volatility.

For now, the​​ main activities of the​​​‌ STACK team on this​ challenge is on the​‌ security and confidentiality axis​​ (PhD thesis of Lomig​​​‌ Piette).

Pulse (Inria-Qarnot Computing)​

Participants: Adrien Lebre.​‌

The joint challenge between​​ Inria and Qarnot computing​​​‌ is called PULSE, for​ "PUshing Low-carbon Services towards​‌ the Edge". It aims​​ to develop and promote​​​‌ best practices in geo-repaired​ hardware and software infrastructures​‌ for more environmentally friendly​​ intensive computing.

The challenge​​​‌ is structured around two​ complementary research axes to​‌ address this technological and​​ environmental issue:

Axis 1:​​​‌ "Holistic analysis of the​ environmental impact of intensive​‌ computing".

Axis 2: "Implementing​​ more virtuous edge services"​​​‌

The STACK group is​ mainly involved in the​‌ second axis, addressing the​​ challenges related to data​​​‌ management.

10.4.6 Inria transfer​ projects (ADT)

CoAnsible

Participants:​‌ Hélène Coullon, Baptiste​​ Jonglez, Sidi-Mohammed Kaddour​​​‌.

The transfer actions​ CoAnsible is a two​‌ year project to develop​​ an Ansible extension able​​​‌ to use the Concerto​ language 56 as an​‌ efficient backend tasks coordinator​​ to handle the lifecycle​​​‌ of infrastructure resources. A​ 24-months software engineer is​‌ hired starting in December​​ 2024.

10.5 Regional initiatives​​​‌

10.5.1 A Cloud-Edge-IoT continuum​ site for SLICES-FR

Participants:​‌ Remous-Aris Koutsiamanis, Jean-Marc​​ Menaud, Baptiste Jonglez​​​‌.

The SLICES-FR site​ in Nantes, comprising the​‌ STACK Research team and​​ the LS2N and IETR​​​‌ UMRs, is advancing the​ development of a Cloud-Edge-IoT​‌ continuum-supporting site. This infrastructure​​ envisions to enable seamless,​​​‌ low-latency experimentation across the​ continuum, contributing a regional​‌ site to the national​​ SLICES-FR and European SLICES-RI​​​‌ platforms.

The current efforts​ are focused on integrating​‌ new hardware and collaborating​​ within the architecture and​​​‌ technical teams to align​ with the project's vision,​‌ funded by multiple national​​ projects for funding hardware​​​‌ acquisition and providing engineering​ resources for management, integration,​‌ and operations, ensuring the​​ sites's robustness and readiness​​​‌ for experimentation.

11 Dissemination​

11.1 Promoting scientific activities​‌

11.1.1 Scientific events: organisation​​

  • Hélène Coullon : Program​​​‌ co-chair of UCC 2025,​ workshops co-chair of CCGrid​‌ 2025.
  • Guillaume Rosinosky :​​ Workshop co-chair of UCC/BDCAT​​ 2025, INSPIRE 2025 workshop​​​‌ co-located with UCC/BDCAT 2025.‌
  • Kandaraj Piamrat : Program‌​‌ co-chair of BDCAT 2025.​​
  • Remous-Aris Koutsiamanis : Tutorial​​​‌ co-chair of UCC 2025.‌
General chair, scientific chair,‌​‌ steering committees
  • Daniel Balouek:​​ General co-chair of UCC/BDCAT​​​‌ 2025, QUICK 2025, workshop‌ co-located with CCGrid, publicity‌​‌ chair of IC2E 2025.​​
Member of the steering​​​‌ committees

Adrien Lebre IEEE‌ ICFEC (International Conference on‌​‌ Fod and Edge Computing)​​

11.1.2 Scientific events: selection​​​‌

Chair of conference program‌ committees
  • Carlos Gonzalez: MedComNet‌​‌ 2025, Session 5: Special​​ Topics in Networked Applications​​​‌
Member of the conference‌ program committees
  • Daniel Balouek‌​‌ : DAIS 2025, IC2E​​ 2025, UCC 2025, HPDC​​​‌ 2025, PAISE 2025, CCGRID‌ 2025, SAC 2025.
  • Thomas‌​‌ Ledoux : IEEE Cloud​​ 2025, UCC 2025.
  • Adrien​​​‌ Lebre : ICFEC 2025,‌ UCC 2025.
  • Kandaraj Piamrat‌​‌ : IEEE CCNC 2025,​​ IEEE ICCE 2025, IEEE​​​‌ ICC 2025, IEEE WiMob‌ 2025, IEEE VCC 2025,‌​‌ IEEE Globecom 2025 (SAC-CECN,​​ SAC-BD), IEEE BDCAT 2025.​​​‌
  • Guillaume Rosinosky : Dais‌ 2025, Middleware 2025.
  • Remous-Aris‌​‌ Koutsiamanis : UCC 2025,​​ IEEE CSCN 2025, IEEE​​​‌ Globecom 2025 CISS, CoRes‌ 2025.
  • Carlos Gonzalez :‌​‌ UCC/BDCAT2025.
  • Mario Südholt :​​ ICISSP 2022-25

11.1.3 Journal​​​‌

Member of the editorial‌ boards
  • Mario Südholt is‌​‌ a member of the​​ advisory board of ”The​​​‌ Programming Journal”
Reviewer -‌ reviewing activities
  • Kandaraj Piamrat‌​‌ : IEEE Transactions on​​ Network and Service Management​​​‌ (TNSM), Future Generation Computer‌ Systems (FGCS), Elsevier Computer‌​‌ Communication (COMCOM)
  • Carlos Gonzalez:​​ IEEE Access, IEEE Sensors​​​‌ Letters, Future Internet, Sensors,‌ Electronics, International Journal of‌​‌ Communication Systems
  • Daniel Balouek:​​ IEEE Internet Computing, Future​​​‌ Generation Computer Systems
  • Remous-Aris‌ Koutsiamanis : MDPI Machine‌​‌ Learning and Knowledge Extraction,​​ Elsevier Expert Systems With​​​‌ Applications, Elsevier Ad Hoc‌ Networks, IEEE Transactions on‌​‌ Network and Service Management​​
  • Guillaume Rosinosky : International​​​‌ Journal of Information Security,‌ Future Generation Computer Systems‌​‌

11.1.4 Invited talks

  • Hélène​​ Coullon : Invited "wide-"speaker​​​‌ at DisCoTeC 2025 (20th‌ International Federated Conference on‌​‌ Distributed Computing Techniques). Challenges​​ in Infrastructure-as-Code: efficiency, decentralization,​​​‌ and formalization.
  • Baptiste Jonglez:‌ Invited tutorials on EnOSlib‌​‌ at DAIS 2025 and​​ the SLICES-FR summer school​​​‌ 2025.
  • Adrien Lebre: Invited‌ talk on “Le programme‌​‌ Systeme, Réseau et Cloud​​ (SRC) et la continuité​​​‌ numérique”, academies des technologies,‌ inter agences du numérique”Academie‌​‌ des technologies, avril 2025.​​
  • Adrien Lebre: Invited talk​​​‌ on “Digital Computing Continuum,‌ the shared vision of‌​‌ two French institutes: Inria​​ and CEA”, DG Connect​​​‌ (Future Network), DG Connect,‌ Brussels, June 2025.
  • Adrien‌​‌ Lebre: Invited talk on​​ “La continuité de la​​​‌ connectivité : le projet‌ inter agences du numérique”‌​‌ CEA workshop, Dec 2025.​​
  • Mario Südholt: Invited talk​​​‌ on the declarative definition‌ and effective implementation of‌​‌ transfer learning algorithms at​​ the Symposium on Systems​​​‌ Science at University of‌ Monastir, Tunesia, 2 Jul.‌​‌ 2025

11.1.5 Scientific expertise​​

  • Remous-Aris Koutsiamanis : SLICES-FR​​​‌ Summer School 2025: Member‌ of the Program committee,‌​‌ Session chair of session​​ "L'expérimentation en question entre​​​‌ complexité et multiculturalisme", and‌ Panel member of session‌​‌ "Construisons ensemble Slices-FR"
  • Adrien​​ Lebre : French representative​​​‌ for the discussion on‌ the possibility of a‌​‌ new European IPCEI related​​​‌ to the computing continuum,​ April 2025.
  • Mario Südholt​‌ : Expert participating in​​ the selection process of​​​‌ the Tiburtius-Preis of the​ best MSc-thesis of all​‌ universities of the Bundesland​​ Berlin, Germany.

11.1.6 Research​​​‌ administration

  • Kandaraj Piamrat is​ the international coordinator of​‌ the LS2N (Laboratoire des​​ sciences du numérique de​​​‌ Nantes).
  • Hélène Coullon is​ the vice-president of the​‌ ACM SigOps France
  • Hélène​​ Coullon is co-chair of​​​‌ the working group YODA​ (trustworthY and Optimal Dynamic​‌ Adaptation) in the national​​ GDR GPL (software engineering​​​‌ and languages)
  • Remous-Aris Koutsiamanis​ is co-chair of the​‌ IETF ROLL (Routing Over​​ Low power and Lossy​​​‌ networks) Working Group
  • Remous-Aris​ Koutsiamanis : Member of​‌ the executive committee of​​ the Grid’5000 GIS (Groupement​​​‌ d’intérêt scientifique), Member of​ the SLICES-FR temporary executive​‌ board, Member of the​​ SLICES-FR Architects committee, Member​​​‌ of the SLICES-FR Site​ managers, Scientific representative of​‌ IMT to SLICES.
  • Adrien​​ Lebre : Member of​​​‌ the executive committee of​ the Grid’5000 GIS (Groupement​‌ d’intérêt scientifique), Member of​​ the SLICES-FR temporary executive​​​‌ board, Scientific French representative​ of the SLICES ISB,​‌ Co-director of the <I/O>​​ Lab, a joint lab​​​‌ between Inria and Orange​ Labs, co-leader the French​‌ CLOUD PEPR, and since​​ December 2023, head of​​​‌ the Cloud, Network and​ System program at the​‌ French agency "Algorithmes, Logiciels​​ et Usages" operated by​​​‌ Inria.

11.2 Teaching -​ Supervision - Juries -​‌ Educational and pedagogical outreach​​

As a team mainly​​​‌ composed of Associate Prof.​ and Prof., the amount​‌ of teaching activities is​​ significant (around 150hours per​​​‌ person). We present here​ only the management activities.​‌

  • Hélène Coullon : responsible​​ for the LOGIN training​​​‌ in computer science at​ IMT Atlantique.
  • Guillaume Rosinosky​‌ : head of the​​ apprenticeship program in Software​​​‌ Engineering FIL. This​ 3-year program leads to​‌ the award of a​​ Master degree in Software​​​‌ Engineering from the IMT​ Atlantique.
  • Thomas Ledoux: member​‌ of the teaching committee​​ at IMT Atlantique
  • Thomas​​​‌ Ledoux: co-pilot of the​ core curriculum of the​‌ 1st year of the​​ IT program at IMT​​​‌ Atlantique
  • Thomas Ledoux :​ head of the Filière​‌ informatique nantaise since Sept.​​ 2020. This entity, created​​​‌ by Nantes University, Centrale​ Nantes and IMT Atlantique,​‌ aims to bring together​​ the main players in​​​‌ Computer Science training in​ Nantes to ensure a​‌ coherent and ambitious training​​ offer that meets the​​​‌ present and future challenges​ of Computer Science. It​‌ is organized around a​​ Council made up of​​​‌ representatives from the academic​ and socio-economic worlds.
  • Thomas​‌ Ledoux : member of​​ the board of directors​​​‌ of Talents du numérique​.
  • Jacques Noyé :​‌ deputy head of the​​ Automation, Production and Computer​​​‌ Sciences Department of IMT​ Atlantique.
  • Mario Südholt :​‌ representative for MSc-level and​​ PhD-level studies of the​​​‌ API department of IMT​ Atlantique.
  • Mario Südholt :​‌ coordinator of the international​​ PhD program SEED (see​​​‌ 10).
  • Kandaraj Piamrat​ : elected member of​‌ scientific council at Faculty​​ of Sciences and Techniques,​​​‌ Nantes University

11.2.1 Supervision​

  • PhD in progress: Olivia​‌ Proust, Towards formally verified​​ configuration management languages, Sept.​​ 2025 - Aug. 2028,​​​‌ Director: Hélène Coullon
  • PhD‌ in progress: Simon Artus,‌​‌ Intelligent and dynamic adaptation​​ of a Cloud2IoT service​​​‌ infrastructure, May 2025 -‌ April 2028, Director: Hélène‌​‌ Coullon
  • PhD in progress:​​ Nathan Rabier, Handling dynamic​​​‌ constraints and deadlines in‌ distributed software reconfiguration -‌​‌ Application to power transmission,​​ Nov. 2025 - Oct.​​​‌ 2028, Director: Hélène Coullon‌
  • PhD in progress: Lucien‌​‌ Astié, Resilience of jointly​​ managed digital service infrastructures,​​​‌ Oct. 2025 - Sept.‌ 2028, Director: Hélène Coullon‌​‌ , Advisor: Baptiste Jonglez​​
  • PhD defended: Duc Thinh​​​‌ Ngo, Dynamic graph learning‌ algorithms for the digital‌​‌ twin in edge-cloud continuum,​​ Dec. 2022 - Nov.​​​‌ 2025, Director: A. Lebre,‌ Advisor: K. Piamrat
  • PhD‌​‌ in progress: Houssem Jmal,​​ Federated Learning for Enhancing​​​‌ Security and Privacy of‌ Decentralized and Distributed Systems,‌​‌ Apr. 2024 - Mar.​​ 2027, Director: JM. Menaud,​​​‌ Advisor: K. Piamrat
  • PhD‌ in progress: Martin Molli,‌​‌ Decision models for the​​ Edge-Cloud Computing Continuum, Nov.​​​‌ 2024 - Sept. 2027,‌ Director: T. Ledoux, Advisor:‌​‌ D. Balouek
  • PhD in​​ progress: Lylian Siffre, Impacts​​​‌ and Uses of Local-First‌ Software for Energy Optimization‌​‌ of IT Services, Nov.​​ 2024 - Sept. 2027,​​​‌ Director: T. Ledoux
  • PhD‌ defended: Antoine Omond, Safe,‌​‌ efficient and low-energy self-adaptation​​ for Cyber Physical Systems​​​‌ - Application to a‌ scientific observatory in the‌​‌ Arctic tundra, Dec. 2021​​ - May 2025, Director:​​​‌ Hélène Coullon
  • PhD change‌ of team/direction: Tengfei An,‌​‌ Modeling and Studying Self-Stabilization​​ within Kubernetes, Oct. 2023​​​‌ - Sept. 2026, Previous‌ Director: A. Lebre, Previous‌​‌ Advisor: H. Coullon, J.​​ Noyé, New Laboratory: LIP6​​​‌
  • PhD in progress: Abdou‌ Seck, Parallel transfer service‌​‌ for the exchange of​​ large volumes of data​​​‌ between Datacenters, Jun. 2022‌ - Jun. 2025, Director:‌​‌ J.-M. Menaud, Noel De​​ Palma, Advisor: R.-A. Koutsiamanis.​​​‌
  • PhD in progress: Samia‌ Boutalbi, Secure deployment of‌​‌ microsservices in a shared​​ RAN/MEC Cloud environment, Jan​​​‌ 2022 -, Director: M.‌ Südholt, M. Dammak, Advisor:‌​‌ R.-A. Koutsiamanis.
  • PhD in​​ progress: Mohammed Abdrrahim Lahmar,​​​‌ Contracts for Distributed ML-Intensive‌ Systems, Oct. 2024 -‌​‌ Oct. 2027, Directors: M.​​ Südholt, Coen De Rover​​​‌ (VU Brussel), Advisor: R.-A.‌ Koutsiamanis.
  • PhD in progress:‌​‌ Irina Samus, Energy-aware actor-based​​ distributed programming, Oct. 25​​​‌ -, Directors: M. Südholt,‌ Jens Nicolay (VU Brussel).‌​‌
  • PhD in progress: Mouheb​​ Jemai, “: Orchestration intelligente​​​‌ et scalable pour garantir‌ les performances et la‌​‌ fiabilité des systèmes Cloud​​ Native”, April 2025 -​​​‌ March 2028, Director: A.‌ Lebre, Advisor: F. Baligand‌​‌ (CEA)
  • PhD defended: Divi​​ Delacour, Architecture and security​​​‌ of cooperative autonomous systems,‌ Jan. 2022 - Jan.‌​‌ 2025, Director: M. Südholt,​​ Marc Lacoste (Orange).
  • PhD​​​‌ defended: Simon Brocard, Premières‌ analyses génomiques de la‌​‌ transplantation pulmonaire, Jan. 22​​ - Dec. 25, Director:​​​‌ M. Südholt, Sophie Limou‌ (Centrale Nantes), Adrian Tissot‌​‌ (CHU Nantes).
  • PhD in​​ progress: Cherif Si Mohammed,​​​‌ Eco-responsible management of data‌ storage, Dec. 2023, Dec.‌​‌ 2026, Director: A. Lebre,​​ Advisor: A. Van Kempen​​​‌ (Qarnot Computing)
  • PhD defended:‌ Hiba Awad, A Model-based‌​‌ Approach for Multi-Scale and​​ Dynamic Distributed Systems, Nov.​​​‌ 2021 - Mar. 2025.‌ Director: T. Ledoux.
  • PhD‌​‌ in progress: Wedwang Romial​​​‌ Menra, Deployment and updating​ electronic equipment, Nov. 2024​‌ - Nov. 2027. Director:​​ J.-M. Menaud
  • PhD in​​​‌ progress: Severin Bradley Anzie,​ Decentralised and sustainable optimisation​‌ of inter-centre data transfers​​ via the QUIC protocol,​​​‌ May 2025, Director: J.-M.​ Menaud, Advisor: R.-A. Koutsiamanis.​‌
  • PhD in progress: Christophe​​ Dion, (industrial “Cifre” thesis​​​‌ with Orange, since Dec.​ 2024), Director: A. Lebre​‌
  • Celeste Precil Guimapi Guefack,​​ Efficacité Énergétique des applications​​​‌ dans le cloud,Jan. 2025,​ Director: J.-M. Menaud, Advisor:​‌ R.-A. Koutsiamanis.
  • PhD in​​ progress: Aymene Boucha, Distributed​​​‌ Machine Learning for the​ Cloud–Edge–IoT (CEI) continuum, Sep.​‌ 2025 - Aug. 2028.​​ Director: M. Südholt, Advisor:​​​‌ K. Piamrat, C. Gonzalez.​
  • PhD in progress: Gaëtan​‌ Plisson, Transparent service continuity​​ for distributed applications on​​​‌ the edge-cloud continuum, Dec.​ 2025 - Nov. 2028​‌ Director: A. Lebre, Advisor:​​ G. Rosinosky, D. Balouek​​​‌
  • PhD in progress: Lomig​ Piette, Privacy-Enabled AI Job​‌ Execution on Heterogeneous Consumer​​ Hardware Architectures, Oct. 2025​​​‌ - Sept. 2028 Director:​ M. Südholt, Advisor: G.​‌ Rosinosky
  • PhD in progress:​​ Ahmed RJIBA, Flexible and​​​‌ Decentralised Application Orchestration in​ Fog Environments, Nov. 2025,​‌ Director: N. Parlavantzas (INRIA​​ MAGELLAN), Advisor: R.-A. Koutsiamanis.​​​‌
  • Post-doc: Anas Mokhtari, A​ Holistic Approach for Designing​‌ Carbon-Aware and Energy-Aware Cloud​​ applications, Sept. 2023 -​​​‌ Jan. 2025, Advisor: T.​ Ledoux, B. Jonglez.
  • Post-doc:​‌ Sidna Jeddou, in Data​​ Management and Network/Computing Continuum,​​​‌ Sep. 2025 - Aug.​ 2026, Advisor: A. Lebre,​‌ C. Gonzalez.

11.2.2 Juries​​

  • Carlos Gonzalez was a​​​‌ jury member of the​ PhD of Abdulkadir Dauda,​‌ Université de Reims Champagne​​ Ardenne - A Secure​​​‌ Edge Gateway for IoT:​ An Adaptive Approach for​‌ Multi-Application and Multi-Protocol Integration,​​ oct. 2025.
  • Remous-Aris Koutsiamanis​​​‌ was a jury member​ of the PhD of​‌ Nikolaos Pavlidis, Democritus University​​ of Thrace, Greece -​​​‌ Optimizing Responsible Decentralized Machine​ Learning Methods for AI​‌ Applications, 3 Nov. 2025.​​
  • Adrien Lebre was a​​​‌ reviewer of the PhD​ defense of Lise Jolicoeur​‌ at University of Bordeaux​​ - Towards secure cluster​​​‌ architectures for HPC workflowse,​ Library, Toolbox, and Evaluation,​‌ Dec. 2025.
  • Thomas Ledoux​​ was a reviewer of​​​‌ the PhD defense of​ Imane Taleb at Rochelle​‌ Université - ”Analyse de​​ l'efficacité énergétique d'applications basées​​​‌ sur des microservices”', june​ 2025.
  • Thomas Ledoux was​‌ a reviewer of the​​ PhD defense of Hugo​​​‌ Montfleur at University of​ Lille - ”Concern-Oriented MicroService​‌ Architecture: Language, Library, Toolbox,​​ and Evaluation”', nov. 2025.​​​‌
  • Kandaraj Piamrat was a​ reviewer of the PhD​‌ defense of Hugo De​​ Oliveira at Université Paris-Saclay​​​‌ - "Reinforcement Learning and​ Federated Learning-based Multi-Band Assignment​‌ for IoT Short Packet​​ Communications", Sept. 2025.
  • Kandaraj​​​‌ Piamrat was a reviewer​ of the PhD defense​‌ of Johann HUGON at​​ ENS Lyon - "Pipelines​​​‌ d'extraction de métriques pour​ la supervision du trafic​‌ réseau sous contraintes systèmes",​​ Dec. 2025.
  • Kandaraj Piamrat​​​‌ was a reviewer of​ the PhD defense of​‌ Javier ERREA MORENO at​​ EUROCOM - "Reinforcement Learning​​​‌ and Federated Learning-based Multi-Band​ Assignment for IoT Short​‌ Packet Communications", Dec. 2025.​​
  • Kandaraj Piamrat was an​​​‌ examiner of the PhD​ defense of Bouchra FAKHER​‌ at Université de Haute-Alsace​​ - "Enhanced Federated Learning​​ for Intelligent Energy Management​​​‌ in IoT- Enabled Smart‌ Buildings", July 2025.
  • Kandaraj‌​‌ Piamrat was an examiner​​ of the PhD defense​​​‌ of Ahcene BOUMHAND at‌ University of Rennes -‌​‌ "Network traffic classification for​​ identifying multi-activity situationsin home​​​‌ environments.", Nov. 2025.
  • Kandaraj‌ Piamrat was an examiner‌​‌ of the PhD defense​​ of Abdelmounaim BOUROUDI at​​​‌ University of Rennes -‌ "Apprentissage par renforcement profond‌​‌ multi-agents pour l'allocation et​​ la planification des ressources​​​‌ en 6G", Nov. 2025.‌
  • Mario Südholt was a‌​‌ reviewer of the PhD​​ defense of Mathis Manthe​​​‌ at INSA-Lyon - "Federated‌ learning in neuroimage segmentation",‌​‌ Jun. 2025.
  • Mario Südholt​​ was an examiner of​​​‌ the PhD defense of‌ Hamza KCHOCK at University‌​‌ of Versailles, Saint Quentin​​ - "Edge AI and​​​‌ 5G Network as Enablers‌ for Immersive Virtual Try-On‌​‌ Applications", Dec. 2025.

11.3​​ Popularization

11.3.1 Participation in​​​‌ Live events

Thomas Ledoux‌ was a speaker at‌​‌ the Techno conference ”Numérique​​ Responsable : Le Déclic​​​‌ pour un Futur Durable”'‌ organized by the Pole‌​‌ Images & Réseaux (Nantes,​​ 13/01/2025)

12 Scientific production​​​‌

12.1 Major publications

12.2 Publications​​​‌ of the year

International‌ journals

Invited conferences​​​‌

International peer-reviewed conferences​

Conferences without proceedings

Doctoral dissertations and habilitation​​​‌ theses

  • 33 thesisH.‌Hiba Awad. Quality‌​‌ of service assurance before​​ deployment of fog systems​​​‌ with model-based engineering and‌ DevOps.Ecole nationale‌​‌ supérieure Mines-Télécom AtlantiqueMarch​​ 2025HALback to​​​‌ textback to text‌
  • 34 thesisH.Hélène‌​‌ Coullon. Efficient Reconfigurations​​ with Programmable Life Cycles:​​​‌ Contributions to Safety, Declarativity,‌ and Decentralization.Nantes‌​‌ UniversitéApril 2025HAL​​back to text
  • 35​​​‌ thesisD.Divi De‌ Lacour. Architecture and‌​‌ security for cooperative autonomous​​ systems.Ecole nationale​​​‌ supérieure Mines-Télécom AtlantiqueJune‌ 2025HALback to‌​‌ text
  • 36 thesisD.-T.​​Duc-Thinh Ngo. Dynamic​​​‌ graph learning algorithms for‌ digital twins of network‌​‌ architectures.Ecole nationale​​ supérieure Mines-Télécom AtlantiqueNovember​​​‌ 2025HALback to‌ text
  • 37 thesisA.‌​‌Antoine Omond. Study​​ of the energy consumption​​​‌ and duration of a‌ cyber-physical system reconfiguration in‌​‌ the Arctic tundra :​​ from experiments on real​​​‌ infrastructure to extensive simulations‌.Ecole nationale supérieure‌​‌ Mines-Télécom Atlantique; Universitetet i​​ TromsøMay 2025HAL​​​‌back to text
  • 38‌ thesisK.Kandaraj Piamrat‌​‌. From network management​​ towards network analytics: a​​​‌ decade journey of research‌ study.Nantes Université‌​‌February 2025HALback​​ to text

Reports &​​​‌ preprints

Other scientific​​ publications

12.3 Cited publications​

  • 45 inproceedingsN.Neil​‌ Ayeb, É.Éric​​ Rutten, S.Sébastien​​​‌ Bolle, T.Thierry​ Coupaye and M.Marc​‌ Douet. Coordinated autonomic​​ loops for target identification,​​​‌ load and error-aware Device​ Management for the IoT​‌.Proceedings of the​​ 2020 Federated Conference on​​​‌ Computer Science and Information​ Systems, FedCSIS 2020, Sofia,​‌ Bulgaria, September 6-9, 2020​​21Annals of Computer​​​‌ Science and Information Systems​2020, 491--500URL:​‌ https://doi.org/10.15439/2020F154DOIback to​​ text
  • 46 inproceedingsN.​​​‌Neil Ayeb, É.​Éric Rutten, S.​‌Sébastien Bolle, T.​​Thierry Coupaye and M.​​​‌Marc Douet. Towards​ an Autonomic and Distributed​‌ Device Management for the​​ Internet of Things.​​​‌IEEE 4th International Workshops​ on Foundations and Applications​‌ of Self* Systems, FAS*W@SASO/ICCAC​​ 2019, Umea, Sweden, June​​​‌ 16-20, 2019IEEE2019​, 246--248URL: https://doi.org/10.1109/FAS-W.2019.00065​‌DOIback to text​​
  • 47 inproceedingsN.Noorani​​​‌ Bakerally, C.Cyrille​ Bareau, F.Fabrice​‌ Blache, S.Sébastien​​ Bolle, C.Christelle​​​‌ Ecrepont, P.Pauline​ Folz, N.Nathalie​‌ Hernandez, T.Thierry​​ Monteil, G.Gilles​​​‌ Privat and F.Fano​ Ramparany. Semi-automatic RDFization​‌ Using Automatically Generated Mappings​​.The Semantic Web:​​​‌ ESWC 2020 Satellite Events​ - ESWC 2020 Satellite​‌ Events, Heraklion, Crete, Greece,​​ May 31 - June​​​‌ 4, 2020, Revised Selected​ Papers12124Lecture Notes​‌ in Computer ScienceSpringer​​2020, 25--31URL:​​​‌ https://doi.org/10.1007/978-3-030-62327-2_5DOIback to​ text
  • 48 inproceedingsD.​‌Daniel Balouek, A.​​ C.Alexandra Carpen Amarie​​​‌, G.Ghislain Charrier​, F.Frédéric Desprez​‌, E.Emmanuel Jeannot​​, E.Emmanuel Jeanvoine​​​‌, A.Adrien Lèbre​, D.David Margery​‌, N.Nicolas Niclausse​​, L.Lucas Nussbaum​​​‌ and others. Adding​ virtualization capabilities to the​‌ Grid’5000 testbed.International​​ Conference on Cloud Computing​​​‌ and Services ScienceSpringer​2012, 3--20back​‌ to text
  • 49 inproceedings​​M.Maxime Belair,​​​‌ S.Sylvie Laniepce and​ J.-M.Jean-Marc Menaud.​‌ SNAPPY: Programmable Kernel-Level Policies​​ for Containers.SAC​​​‌ 2021: 36th ACM/SIGAPP Symposium​ On Applied ComputingGwangju​‌ / Virtual, South Korea​​March 2021, 1636--1645​​​‌HALDOIback to​ text
  • 50 articleG.​‌Gordon Blair, T.​​Thierry Coupaye and J.-B.​​​‌Jean-Bernard Stefani. Component-based​ architecture: the Fractal initiative​‌.Annals of telecommunications​​641February 2009​​​‌, 1--4URL: https://doi.org/10.1007/s12243-009-0086-1​DOIback to text​‌
  • 51 unpublishedS.Samia​​ Boutalbi. Secure deployment​​​‌ of micro-services in a​ shared Cloud RAN/MEC computing​‌ environment.12 2025​​, PhD defense planned​​​‌ for March/April 2026back​ to text
  • 52 article​‌S.Simon Brocard,​​ M.Martin Morin,​​​‌ N.Nayane dos Santos​ Brito Silva, B.​‌Benjamin Renaud-Picard, B.​​Benjamin Coiffard, X.​​​‌Xavier Demant, L.​Lo\"ic Falque, J.​‌Jérome Le Pavec,​​ A.Antoine Roux,​​ T.Thomas Villeneuve,​​​‌ C.Christiane Knoop,‌ J.-F.Jean-François Mornex,‌​‌ M.Mathilde Salpin,​​ V.Véronique Boussaud,​​​‌ O.Olivia Rousseau,‌ V.Vincent Mauduit,‌​‌ A.Axelle Durand,​​ A.Antoine Magnan,​​​‌ P.-A.Pierre-Antoine Gourraud,‌ N.Nicolas Vince,‌​‌ M.Mario Südholt,​​ A.Adrien Tissot and​​​‌ S.Sophie Limou.‌ Description and first insights‌​‌ on a large genomic​​ biobank of lung transplantation​​​‌.European Journal of‌ Human GeneticsAugust 2024‌​‌, 1-8HALDOI​​back to text
  • 53​​​‌ inproceedingsH.Hugo Bruneliere‌, Z.Zakarea Al-Shara‌​‌, F.Frederico Alvares​​, J.Jonathan Lejeune​​​‌ and T.Thomas Ledoux‌. A Model-based Architecture‌​‌ for Autonomic and Heterogeneous​​ Cloud Systems.CLOSER​​​‌ 2018 - 8h International‌ Conference on Cloud Computing‌​‌ and Services Science1​​Best Paper AwardFunchal,​​​‌ PortugalMarch 2018,‌ 201-212HALDOIback‌​‌ to text
  • 54 article​​A.Amaury Bruniaux,​​​‌ R.-A.Remous-Aris Koutsiamanis,‌ G. Z.Georgios Z‌​‌ Papadopoulos and N.Nicolas​​ Montavont. Defragmenting the​​​‌ 6LoWPAN Fragmentation Landscape: A‌ Performance Evaluation.Sensors‌​‌2152021,​​ 1711back to text​​​‌
  • 55 inproceedingsM.Maverick‌ Chardet, H.Hélène‌​‌ Coullon, D.Dimitri​​ Pertin and C.Christian​​​‌ Pérez. Madeus: a‌ formal deployment model.‌​‌2018 International Conference on​​ High Performance Computing &​​​‌ Simulation (HPCS)IEEE2018‌, 724--731back to‌​‌ text
  • 56 articleM.​​Maverick Chardet, H.​​​‌Hélène Coullon and S.‌Simon Robillard. Toward‌​‌ Safe and Efficient Reconfiguration​​ with Concerto.Science​​​‌ of Computer Programming203‌March 2021, 1-31‌​‌HALDOIback to​​ textback to text​​​‌back to textback‌ to text
  • 57 article‌​‌M.Maximilien Charlier,​​ R.-A.Remous-Aris Koutsiamanis and​​​‌ B.Bruno Quoitin.‌ Scheduling UWB Ranging and‌​‌ Backbone Communications in a​​ Pure Wireless Indoor Positioning​​​‌ System.IoT3‌12022, 219--258‌​‌back to textback​​ to text
  • 58 article​​​‌R.-A.Ronan-Alexandre Cherrueau,‌ M.Marie Delavergne,‌​‌ A.Alexandre van Kempen​​, A.Adrien Lebre​​​‌, D.Dimitri Pertin‌, J.Javier Rojas‌​‌ Balderrama, A.Anthony​​ Simonet and M.Matthieu​​​‌ Simonin. EnosLib: A‌ Library for Experiment-Driven Research‌​‌ in Distributed Computing.​​IEEE Transactions on Parallel​​​‌ and Distributed Systems33‌6June 2022,‌​‌ 1464-1477HALDOIback​​ to text
  • 59 inproceedings​​​‌R.-A.Ronan-Alexandre Cherrueau,‌ D.Dimitri Pertin,‌​‌ A.Anthony Simonet,​​ A.Adrien Lebre and​​​‌ M.Matthieu Simonin.‌ Toward a Holistic Framework‌​‌ for Conducting Scientific Evaluations​​ of OpenStack.Proceedings​​​‌ of the 17th IEEE/ACM‌ International Symposium on Cluster,‌​‌ Cloud and Grid Computing​​IEEE Press2017,​​​‌ 544--548back to text‌
  • 60 inproceedingsM.Manuel‌​‌ Clavel, F.Fransisco​​ Durán, S.Steven​​​‌ Eker, P.Patrick‌ Lincoln, N.Narciso‌​‌ Martí-Oliet, J.José​​ Meseguer and J. F.​​​‌Jose F. Quesada.‌ The Maude System.‌​‌Proceedings of the 10th​​ International Conference on Rewriting​​​‌ Techniques and ApplicationsRtA‌ '99Berlin, HeidelbergSpringer-Verlag‌​‌1999back to text​​​‌
  • 61 articleR. D.​Roberto Di Cosmo,​‌ J.Jacopo Mauro,​​ S.Stefano Zacchiroli and​​​‌ G.Gianluigi Zavattaro.​ Aeolus: A component model​‌ for the cloud.​​Information and Computation239​​​‌Supplement C2014,​ 100--121URL: http://www.sciencedirect.com/science/article/pii/S0890540114001424DOI​‌back to text
  • 62​​ articleH.Hélène Coullon​​​‌, J.Julien Bigot​ and C.Christian Pérez​‌. Extensibility and Composability​​ of a Multi-Stencil Domain​​​‌ Specific Framework.International​ Journal of Parallel Programming​‌November 2017, URL:​​ https://doi.org/10.1007/s10766-017-0539-5DOIback to​​​‌ text
  • 63 articleH.​Hélène Coullon, L.​‌Ludovic Henrio, F.​​Frédéric Loulergue and S.​​​‌Simon Robillard. Component-Based​ Distributed Software Reconfiguration: a​‌ Verification-Oriented Survey.ACM​​ Computing Surveys561​​​‌January 2024, 1-37​HALDOIback to​‌ text
  • 64 inproceedingsT.​​Thierry Coupaye, S.​​​‌Sébastien Bolle, S.​Sylvie Derrien, P.​‌Pauline Folz, P.​​Pierre Meye, G.​​​‌Gilles Privat and P.​ R.Philippe Raipin Parvedy​‌. A graph-based cross-vertical​​ digital twin platform for​​​‌ complex cyber-physical systems.​The Digital TwinIn​‌ press, to appearSpringer,​​ Cham2022back to​​​‌ text
  • 65 inproceedingsT.​Thierry Coupaye and J.-B.​‌Jean-Bernard Stefani. Fractal​​ Component-Based Software Engineering.​​​‌Object-Oriented Technology, ECOOP 2006​ Workshop Reader, ECOOP 2006​‌ Workshops, Nantes, France, July​​ 3-7, 2006, Final Reports​​​‌4379Lecture Notes in​ Computer ScienceSpringer2006​‌, 117--129URL: https://doi.org/10.1007/978-3-540-71774-4_13​​DOIback to text​​​‌
  • 66 articleP.-C.Pierre-Charles​ David, T.Thomas​‌ Ledoux, M.Marc​​ Léger and T.Thierry​​​‌ Coupaye. FPath and​ FScript: Language support for​‌ navigation and reliable reconfiguration​​ of Fractal architectures.​​​‌annals of telecommunications -​ annales des télécommunications64​‌1February 2009,​​ 45--63URL: https://doi.org/10.1007/s12243-008-0073-yDOI​​​‌back to text
  • 67​ inproceedingsF. A.Frederico​‌ Alvares De Oliveira,​​ T.Thomas Ledoux and​​​‌ R.Rémi Sharrock.​ A framework for the​‌ coordination of multiple autonomic​​ managers in cloud environments​​​‌.Self-Adaptive and Self-Organizing​ Systems (SASO), 2013 IEEE​‌ 7th International Conference on​​IEEE2013, 179--188​​​‌back to text
  • 68​ inproceedingsX.Xavier Etchevers​‌, G.Gwen Salaün​​, F.Fabienne Boyer​​​‌, T.Thierry Coupaye​ and N. D.Noël​‌ De Palma. Reliable​​ self-deployment of cloud applications​​​‌.Symposium on Applied​ Computing, SAC 2014, Gyeongju,​‌ Republic of Korea -​​ March 24 - 28,​​​‌ 2014ACM2014,​ 1331--1338URL: https://doi.org/10.1145/2554850.2554951DOI​‌back to text
  • 69​​ articleX.Xavier Etchevers​​​‌, G.Gwen Salaün​, F.Fabienne Boyer​‌, T.Thierry Coupaye​​ and N. D.Noel​​​‌ De Palma. Reliable​ self-deployment of distributed cloud​‌ applications.Softw. Pract.​​ Exp.4712017​​​‌, 3--20URL: https://doi.org/10.1002/spe.2400​DOIback to text​‌
  • 70 articleO.Olivier​​ Flauzac, C.Carlos​​​‌ Javier Gonzalez Santamaria,​ F.Florent Nolot and​‌ I.Isaac Woungang.​​ An SDN approach to​​​‌ route massive data flows​ of sensor networks.​‌International Journal of Communication​​ Systems337e4309​​​‌ dac.43092020, e4309​URL: https://onlinelibrary.wiley.com/doi/abs/10.1002/dac.4309DOIback​‌ to text
  • 71 inproceedings​​V.Vaidas Gasiūnas,​​ L.Lucas Satabin,​​​‌ M.Mira Mezini,‌ A.Angel Núñez and‌​‌ J.Jacques Noyé.​​ EScala: Modular Event-Driven Object​​​‌ Interactions in Scala.‌URL: http://doi.acm.org/10.1145/1960275.1960303back to‌​‌ text
  • 72 inproceedingsC.​​Carlos Gonzalez and S.​​​‌ M.Salim Mahamat Charfadine‌. SDN Controllers and‌​‌ ML-Based Anomaly Detection in​​ Embedded Systems: A Comparative​​​‌ Analysis.2023 10th‌ International Conference on Wireless‌​‌ Networks and Mobile Communications​​ (WINCOM)2023, 1-6​​​‌DOIback to text‌
  • 73 inproceedingsC.Cyprien‌​‌ Gottstein, P. R.​​Philippe Raipin Parvedy,​​​‌ M.Michel Hurfin,‌ T.Thomas Hassan and‌​‌ T.Thierry Coupaye.​​ Inverse Space Filling Curve​​​‌ Partitioning Applied to Wide‌ Area Graphs.DMS‌​‌ 2020 - 11th International​​ conference on Database Management​​​‌ SystemsZurich, SwitzerlandNovember‌ 2020, 223-241HAL‌​‌DOIback to text​​
  • 74 articleC.Cyprien​​​‌ Gottstein, P. R.‌Philippe Raipin Parvedy,‌​‌ M.Michel Hurfin,​​ T.Thomas Hassan and​​​‌ T.Thierry Coupaye.‌ Partitioning Wide Area Graphs‌​‌ Using a Space Filling​​ Curve.International Journal​​​‌ of Data Mining &‌ Knowledge Management Process11‌​‌1January 2021,​​ 13-31HALDOIback​​​‌ to text
  • 75 article‌D.David Hauweele,‌​‌ R.-A.Remous-Aris Koutsiamanis,​​ B.Bruno Quoitin and​​​‌ G. Z.Georgios Z‌ Papadopoulos. Thorough performance‌​‌ evaluation & analysis of​​ the 6TiSCH minimal scheduling​​​‌ function (MSF).Journal‌ of Signal Processing Systems‌​‌9412022,​​ 3--25back to text​​​‌back to text
  • 76‌ inproceedingsF.Fabien Hermenier‌​‌, X.Xavier Lorca​​, J.-M.Jean-Marc Menaud​​​‌, G.Gilles Muller‌ and J.Julia Lawall‌​‌. Entropy: a consolidation​​ manager for clusters.​​​‌Proceedings of the 2009‌ ACM SIGPLAN/SIGOPS international conference‌​‌ on Virtual execution environments​​ACM2009, 41--50​​​‌back to text
  • 77‌ articleT. L.Tomas‌​‌ Lagos Jenschke, R.-A.​​Remous-Aris Koutsiamanis, G.​​​‌ Z.Georgios Z Papadopoulos‌ and N.Nicolas Montavont‌​‌. ODeSe: On-Demand Selection​​ for multi-path RPL networks​​​‌.Ad Hoc Networks‌1142021, 102431‌​‌back to textback​​ to text
  • 78 article​​​‌J.J.O. Kephart and‌ D.D.M. Chess.‌​‌ The vision of autonomic​​ computing.Computer36​​​‌12003, 41-50‌DOIback to text‌​‌
  • 79 inproceedingsR.-A.Remous-Aris​​ Koutsiamanis, G. Z.​​​‌Georgios Z Papadopoulos,‌ B.Bruno Quoitin and‌​‌ N.Nicolas Montavont.​​ A Centralized Controller for​​​‌ Reliable and Available Wireless‌ Schedules in Industrial Networks‌​‌.2020 16th International​​ Conference on Mobility, Sensing​​​‌ and Networking (MSN)IEEE‌2020, 1--9back‌​‌ to textback to​​ text
  • 80 articleA.​​​‌Adrien Lebre, J.‌Jonathan Pastor, A.‌​‌Anthony Simonet and M.​​Mario Südholt. Putting​​​‌ the Next 500 VM‌ Placement Algorithms to the‌​‌ Acid Test.IEEE​​ Transactions on Parallel and​​​‌ Distributed Systems2018,‌ URL: https://hal.inria.fr/hal-01816248back to‌​‌ text
  • 81 inproceedingsM.​​Marc Léger, T.​​​‌Thomas Ledoux and T.‌Thierry Coupaye. Reliable‌​‌ Dynamic Reconfigurations in a​​ Reflective Component Model.​​​‌Proceedings of the 13th‌ International Conference on Component-Based‌​‌ Software EngineeringCBSE'10Berlin,​​​‌ HeidelbergPrague, Czech Republic​Springer-Verlag2010, 74--92​‌URL: http://dx.doi.org/10.1007/978-3-642-13238-4_5DOIback​​ to text
  • 82 conference​​​‌J.Jonathan Lejeune,​ F.Frederico Alvares and​‌ T.Thomas Ledoux.​​ Towards a Generic Autonomic​​​‌ Model to Manage Cloud​ Services.Proceedings of​‌ the 7th International Conference​​ on Cloud Computing and​​​‌ Services Science (Best Paper​ Award),INSTICCSciTePress2017​‌, 175-186DOIback​​ to text
  • 83 inproceedings​​​‌L.Loic Letondeur,​ X.Xavier Etchevers,​‌ T.Thierry Coupaye,​​ F.Fabienne Boyer and​​​‌ N. D.Noel De​ Palma. Architectural Model​‌ and Planification Algorithm for​​ the Self-Management of Elastic​​​‌ Cloud Applications.2014​ International Conference on Cloud​‌ and Autonomic Computing, London,​​ United Kingdom, September 8-12,​​​‌ 2014IEEE Computer Society​2014, 172--179URL:​‌ https://doi.org/10.1109/ICCAC.2014.29DOIback to​​ text
  • 84 inproceedingsY.​​​‌Yu Liu, P.​Pauline Folz, S.​‌Shenle Pan, F.​​Fano Ramparany, S.​​​‌Sébastien Bolle, E.​Eric Ballot and T.​‌Thierry Coupaye. Digital​​ Twin-Driven Approach for Smart​​​‌ City Logistics: The Case​ of Freight Parking Management​‌.Advances in Production​​ Management Systems. Artificial Intelligence​​​‌ for Sustainable and Resilient​ Production Systems - IFIP​‌ WG 5.7 International Conference,​​ APMS 2021, Nantes, France,​​​‌ September 5-9, 2021, Proceedings,​ Part IV633IFIP​‌ Advances in Information and​​ Communication TechnologySpringer2021​​​‌, 237--246URL: https://doi.org/10.1007/978-3-030-85910-7_25​DOIback to text​‌
  • 85 inproceedingsA.Antoine​​ Omond, H.Hélène​​​‌ Coullon, I.Issam​ Ra\"is and O.Otto​‌ Anshus. Leveraging Relay​​ Nodes to Deploy and​​​‌ Update Services in a​ CPS with Sleeping Nodes​‌.Proceeding of the​​ 2023 IEEE International Conferences​​​‌ on Internet of Things​ (iThings) and IEEE Green​‌ Computing & Communications (GreenCom)​​ and IEEE Cyber, Physical​​​‌ & Social Computing (CPSCom)​ and IEEE Smart Data​‌ (SmartData) and IEEE Congress​​ on Cybermatics (Cybermatics)December​​​‌ 2023, URL: https://hal.science/hal-04372320​DOIback to text​‌
  • 86 inproceedingsJ.Jolan​​ Philippe, A.Antoine​​​‌ Omond, H.Hélène​ Coullon, C.Charles​‌ Prud'Homme and I.Issam​​ Ra\"is. Fast Choreography​​​‌ of Cross-DevOps Reconfiguration with​ Ballet: A Multi-Site OpenStack​‌ Case Study.SANER​​ 2024: IEEE International Conference​​​‌ on Software Analysis, Evolution​ and ReengineeringRovaniemi, Finland​‌IEEEMarch 2024,​​ 1-11HALDOIback​​​‌ to text
  • 87 inproceedings​G.Gilles Privat,​‌ T.Thierry Coupaye,​​ S.Sébastien Bolle and​​​‌ P.Philippe Raipin Parvedy​. WoT Graph as​‌ Multiscale Digital-Twin for Cyber-Physical​​ Systems-of-Systems.Proc., 2nd​​​‌ W3C Web of Things​ Workshop. Grenoble, France06​‌ 2019back to text​​
  • 88 articleF.Flavien​​​‌ Quesnel, A.Adrien​ Lebre and M.Mario​‌ Südholt. Cooperative and​​ reactive scheduling in large-scale​​​‌ virtualized platforms with DVMS​.Concurrency and Computation:​‌ Practice and Experience25​​122013, 1643--1655​​​‌back to text
  • 89​ inproceedingsF.Fano Ramparany​‌, M.Marceau Thalgott​​, S.Sébastien Bolle​​​‌ and S.Serge Martin​. Exploiting Data Analytics​‌ for Home Automation Services​​.4th IEEE International​​​‌ Conference on Future Internet​ of Things and Cloud,​‌ FiCloud 2016, Vienna, Austria,​​ August 22-24, 2016IEEE​​ Computer Society2016,​​​‌ 228--234URL: https://doi.org/10.1109/FiCloud.2016.40DOI‌back to text
  • 90‌​‌ inproceedingsS.Simon Robillard​​ and H.Hélène Coullon​​​‌. SMT-Based Planning Synthesis‌ for Distributed System Reconfigurations‌​‌.Lecture Notes in​​ Computer ScienceLecture Notes​​​‌ in Computer ScienceMunich,‌ GermanyApril 2022,‌​‌ URL: https://inria.hal.science/hal-03536643DOIback​​ to text
  • 91 inproceedings​​​‌G.Guillaume Rosinosky,‌ D.Donatien Schmitz and‌​‌ E.Etienne Rivière.​​ Streambed: capacity planning for​​​‌ stream processing.Proceedings‌ of the 18th ACM‌​‌ International Conference on Distributed​​ and Event-based Systems2024​​​‌, 90--102back to‌ text
  • 92 articleF.‌​‌ A.Farah A\"it Salaht​​, F.Frédéric Desprez​​​‌ and A.Adrien Lebre‌. An overview of‌​‌ service placement problem in​​ fog and edge computing​​​‌.ACM Computing Surveys‌ (CSUR)5332020‌​‌, 1--35back to​​ text
  • 93 inproceedingsF.​​​‌ A.Farah Ait Salaht‌, F.Frédéric Desprez‌​‌, A.Adrien Lebre​​, C.Charles Prud'Homme​​​‌ and M.Mohamed Abderrahim‌. Service placement in‌​‌ fog computing using constraint​​ programming.2019 IEEE​​​‌ International Conference on Services‌ Computing (SCC)IEEE2019‌​‌, 19--27back to​​ text
  • 94 articleM.​​​‌M. Satyanarayanan. The‌ Emergence of Edge Computing‌​‌.Computer501​​January 2017, 30-39​​​‌DOIback to text‌
  • 95 articleD.Damian‌​‌ Serrano, S.Sara​​ Bouchenak, Y.Yousri​​​‌ Kouki, F. A.‌Frederico Alvares de Oliveira‌​‌ Jr., T.Thomas​​ Ledoux, J.Jonathan​​​‌ Lejeune, J.Julien‌ Sopena, L.Luciana‌​‌ Arantes and P.Pierre​​ Sens. SLA guarantees​​​‌ for cloud services.‌Future Generation Computer Systems‌​‌54Supplement C2016​​, 233--246URL: http://www.sciencedirect.com/science/article/pii/S0167739X15000801​​​‌DOIback to text‌
  • 96 articleW.Weisong‌​‌ Shi, J.Jie​​ Cao, Q.Quan​​​‌ Zhang, Y.Youhuizi‌ Li and L.Lanyu‌​‌ Xu. Edge computing:​​ Vision and challenges.​​​‌IEEE Internet of Things‌ Journal352016‌​‌, 637--646back to​​ text
  • 97 inproceedingsJ.​​​‌ M.Jurgen M. Van‌ Ham, G.Guido‌​‌ Salvaneschi, M.Mira​​ Mezini and J.Jacques​​​‌ Noyé. JEScala: Modular‌ Coordination with Declarative Events‌​‌ and Joins.205--216​​URL: http://doi.acm.org/10.1145/2577080.2577082back to​​​‌ text
  • 98 inproceedingsY.‌Ye Xia, X.‌​‌Xavier Etchevers, L.​​Lo\"ic Letondeur, T.​​​‌Thierry Coupaye and F.‌Frédéric Desprez. Optimizing‌​‌ Cloud Application Scheduling: A​​ Dual-Stage Heuristic Approach.​​​‌ICCCBDA 2024 - 9th‌ International Conference on Cloud‌​‌ Computing and Big Data​​ AnalyticsChengdu, ChinaIEEE​​​‌April 2024, 134-140‌HALDOIback to‌​‌ text
  • 99 inproceedingsY.​​Ye Xia, X.​​​‌Xavier Etchevers, L.‌Loic Letondeur, A.‌​‌Adrien Lebre, T.​​Thierry Coupaye and F.​​​‌Frédéric Desprez. Combining‌ heuristics to optimize and‌​‌ scale the placement of​​ iot applications in the​​​‌ fog.2018 IEEE/ACM‌ 11th International Conference on‌​‌ Utility and Cloud Computing​​ (UCC)IEEE2018,​​​‌ 153--163back to text‌
  • 100 inproceedingsY.Ying‌​‌ Xiong, Y.Yulin​​ Sun, L.Li​​​‌ Xing and Y.Ying‌ Huang. Extend Cloud‌​‌ to Edge with KubeEdge​​​‌.2018 IEEE/ACM Symposium​ on Edge Computing (SEC)​‌2018, 373-377DOI​​back to text