2023Activity reportProject-TeamDIANA

6.1.1 ACQUAmobile

• Name:
  Application for prediCting Quality of User Experience at Internet Access
• Keywords:
  Android, Internet access, Performance measure, Quality of Experience
• Scientific Description:

  ACQUA is an Application for prediCting QUality of Experience (QoE) at Internet Access. It is developed by the Diana team at Inria Sophia Antipolis – Méditerranée and was supported by Inria under the ADT ACQUA grant. The scientific project around ACQUA is supported by Inria Project Lab BetterNet and the French National Project ANR BottleNet. The project also got the approval of Inria COERLE and French CNIL for the part on experimentation with real users. ACQUA presents a new way for the evaluation of the performance of Internet access. Starting from network-level measurements as the ones we often do today (bandwidth, delay, loss rates, jitter, etc), ACQUA targets the estimated Quality of Experience (QoE) related to the different applications of interest to the user without the need to run them (e.g., estimated Skype quality, estimated video streaming quality).

  The ACQUA Android application is supposed to be on one hand the reference application for QoE forecasting and troubleshooting for end users at their Internet access, and on the other hand, the feedback channel that allows end users to report to us (if they are willing) on their experience together with the corresponding network measurements so as to help us calibrating better and more realistic models. For this calibration, we are currently performing extensive, efficient and automatic measurements in the laboratory, we will count on end users to help us completing this dataset with further applications and more realistic network and user conditions.

  ACQUA is mainly meant for end users, but it is also of interest to (mobile) network operators and to content providers to estimate the QoE of their customers and their networks without each time having to run expensive application-level traffic and to involve real users.

• Functional Description:

  An application in ACQUA is a function, or a model, that links the network-level and device-level measurements to the expected Quality of Experience. Supervised machine learning techniques are used to establish such link between measurements both at the network level and the device level, and estimations of the Quality of Experience for different Internet applications. The required data for such learning can be obtained either by controlled experiments as we did in a prior work on Skype and YouTube Quality of Experience, or by soliciting the crowd (i.e. crowdsourcing) for combinations (i.e. tuples) of network- and application-level measurements and corresponding user-level Quality of Experience. Our current work is concentrating on calibrating further models for ACQUA and on using the dataset of ACQUA to understand the performance of mobile networks and user experience in the wild. We refer to the application web site of the project for further details.

  Assessment: Audience = 3, Software Originality = 4, Software Maturity = 3, Evolution and Maintenance = 3, Software Distribution and Licensing = 5.

• URL:
  http://­project.­inria.­fr/­acqua/
• Authors:
  Thierry Spetebroot, Chadi Barakat
• Contact:
  Chadi Barakat

6.1.2 ElectroSmart

• Keywords:
  Crowd-sourcing, UMTS, GSM, Bluetooth, Wi-Fi, 4G, 3G, 2G, Electromagnetic waves, Android, LTE
• Functional Description:

  The Internet and new devices such as smartphones have changed fundamentally the way people communicate, but this technological revolution comes at the price of a higher exposition of the general population to microwave electromagnetic fields (EMF). This exposition is a concern for health agencies and epidemiologists who want to understand the impact of such an exposition on health, for the general public who wants a higher transparency on its exposition and the health hazard it might represent, but also for cellular operators and regulation authorities who want to improve the cellular coverage while limiting the exposition, and for computer scientists who want to better understand the network connectivity in order to optimize communication protocols. Despite the fundamental importance to understand the exposition of the general public to EMF, it is poorly understood because of the formidable difficulty to measure, model, and analyze this exposition.

  The goal of the ElectroSmart project is to develop the instrument, methods, and models to compute the exposition of the general public to microwave electromagnetic fields used by wireless protocols and infrastructures such as Wi-Fi, Bluetooth, or cellular. Using a pluri-disciplinary approach combining crowd-based measurements, in-lab experiments, and modeling using sparse and noisy data, we address challenges such as designing and implementing a measuring instrument leveraging on crowd-based measurements from mobile devices such as smartphones, modeling the exposition of the general public to EMF to compute the most accurate estimation of the exposition, and analyzing the evolution of the exposition to EMF with time. This technological breakthrough will have scientific, technical, and societal applications, notably on public health politics, by providing the scientific community and potential users with a unique measuring instrument, methods, and models to exploit the invaluable data gathered by the instrument. This project was supported by the UCN@Sophia Labex in 2016/2017/2018 (funding the engineer Mondi Ravi), by an Inria ADT (funding the engineer Abdelhakim Akodadi) 2017/2018, by Inria ATT (funding the business developer David Migliacci) in 2017/2018, and by the academy 1 of UCAJedi (funding a Ph.D. student Yanis Boussad) 2017/2021.

  In August 2016, we released the first stable public release of ElectroSmart. On the 30th December 2022 the app has been downloaded more than 3 million times, we have 200 000 active users and a score of 4,5/5 on Google Play. On October 2022, the code has been opened with a BSD 3-Clause license on github and has currently 33 stars and 9 forks.

• URL:
  https://­www-sop.­inria.­fr/­members/­Arnaud.­Legout/­Projects/­electrosmart.­html
• Contact:
  Arnaud Legout
• Participant:
  Arnaud Legout

6.1.3 nepi-ng

• Keywords:
  Wireless network, Experimentation
• Functional Description:

  In the specific context of R2lab, we have created a tool suite for orchestrating network experiments, that for historical reasons we refer to collectively as nepi-ng, for NEPI new generation. An umbrella website is available.

  At this point, nepi-ng has a much smaller scope than its NEPI ancestor used to have, in that it only supports remote control of network experiments over ssh. As a matter of fact, in practice, this is the only access mechanism that we need to have for running experiments on both R2lab, and PlanetLab Europe.

  The design of nepi-ng of course is modular, so that it will be perfectly possible to add other control mechanisms to this core if and when it becomes necessary.

  • asynciojobs:
    • URL: https://­asynciojobs.­readthedocs.­io/
    • Version: asynciojobs v0.17.0
    • Keywords: asynchronous programming, coroutines, orchestration
    • License: CC BY-SA 4.0
    • Type of human computer interaction: Python library
    • OS/Middleware: Linux
    • Required library or software: Python-3.9 / asyncio
    • Programming language: Python
  • apssh:
    • URL: https://­apssh.­readthedocs.­io/
    • Version: apssh v0.24.0
    • Keywords: orchestration, ssh, networking experimentation
    • License: CC BY-SA 4.0
    • Type of human computer interaction: Python library
    • OS/Middleware: Linux
    • Required library or software: Python-3.9 / asyncio
    • Programming language: Python
• URL:
  https://­nepi-ng.­inria.­fr
• Contact:
  Thierry Parmentelat

6.1.4 Distrinet

• Name:
  Distrinet
• Keywords:
  SDN, Emulation, Large-scale Emulators, Network simulator
• Scientific Description:

  Networks have become complex systems that combine various concepts, techniques, and technologies. As a consequence, modelling or simulating them now is extremely complicated and researchers massively resort to prototyping techniques. Mininet is the most popular tool when it comes to evaluate SDN propositions. Mininet allows to emulate SDN networks on a single computer but shows its limitations with resource intensive experiments as the emulating host may become overloaded. To tackle this issue, we propose Distrinet, a distributed implementation of Mininet over multiple hosts, based on LXD/LXC, Ansible, and VXLAN tunnels. Distrinet uses the same API than Mininet, meaning that it is compatible with Mininet programs. It is generic and can deploy experiments on Linux clusters (e.g., Grid’5000), as well as on the Amazon EC2 cloud platform.

  Assessment: A5, SO3, SM2, EM2-down, SDL4

• Functional Description:
  Distrinet is an extension of Mininet that relies on LXC to be distributed in the cloud, and particularly in Amazon. The extension has been designed to be fully compatible with Mininet. As using Distrinet potentially involves the collaboration of multiple machines we focused on guaranteeing the correctness (in a sense that results are trustworthy) of simulations when running on multiple machines. To speedup deployments, loading and unloading operations have been parallelised with asynchronous calls. The pool of machines used for simulations is automatically provisioned thanks to Ansible.
• Release Contributions:
  First release
• URL:
  https://­distrinet-emu.­github.­io
• Publication:
  hal-03000617v1
• Contact:
  Walid Dabbous
• Participants:
  Damien Saucez, Giuseppe Di Lena, Andrea Tomassilli, Frédéric Giroire, Thierry Turletti
• Partner:
  Orange Labs

6.1.5 OMNET-TSN

• Name:
  OMNET-TSN
• Keywords:
  Real time, Network simulator
• Functional Description:

  Time Sensitive Networking (TSN) aims at providing real time capabilities to Ethernet networks. To achieve this goal, the 2018 revision of the IEEE802.1Q standard (i.e., IEEE802.1Q-2018) provides the ability to define transmission selection algorithms for queues in IEE802.1Q Ethernet switches.

  We are developing an IEEE802.1Q-2018 simulation module for OMNeT++. The usual way to implement a simulation module is to abstract the components to be simulated and leverage the simulator functionalities. As a result, simulations can run very efficiently and scale well. The drawback is that the code base is only understandable by well trained developers and hardly understandable by engineers.

  With our implementation we decided to make a direct transpose of the normative documents into the simulator such that anyone that reads the standards can read the code and adapt it if needed. This approach makes simulations runs last much longer than if the code were optimized for simulations but the major advantage is that engineers used to read standards and evaluate appliances provided by third parties can use it with minimal training.

  The simulator code is implemented in C++ in Omnest (the commercial version of OMNeT++). As in most industrial environments it is extremely complex to install non corporate software, we implemented a front-end in javascript with the REACT framework. As a consequence, it is possible to install and run the simulation tool on a dedicated machine/VM/container and to drive simulations from any workstation solely by using a web browser.

• Release Contributions:
   

  • Multiple bug fixes in CBS
  • CBS as a transmission selection algorithm instead of a transmission queue
  • Complete tutorials for users
  • Full code documentation for developper
  • Delete message from QeuingFrame::handleMessage after it has been cloned to reduce memory footprint for large experiments
  • Includes std algorithm library to compile on Linux std::set_intersection
  • Add processing delay simulation
  • Automated testing with docker
• URL:
  https://­github.­com/­dsaucez/
• Contact:
  Damien Saucez

6.2.1 Reproducible research Lab - R2lab

Scientific work around network protocols and related software stacks requires experiments, hence experimental conditions, to be reproducible. This is a particularly challenging requirement in the wireless networking area, where characteristics of wireless channels are known to be variable, unpredictable and hardly controllable.

The R2lab wireless testbed was designed with reproducibility as its central characteristics; it is built around an isolated and anechoic chamber, featuring RF absorbers that prevent radio waves reflections, and a Faraday cage blocking external interferences. R2lab thus provides an ideal environment for running reproducible wireless experiments.

R2lab has been operated since December 2015, in the context of the FIT (Future Internet of Things) Equipment of Excellence project, and as such, it is now federated with the other testbeds that are part of the FIT initiative. As of early 2019, it has been also federated within the Fed4Fire initiative.

Available toolsets, both hardware and software, are mostly stable apart from low noise marginal deployment of new kinds of radio devices, that now encompass 5G and LoRa (from Long Range, a physical proprietary radio communication technique), among others. Our focus at this point of the project is to leverage our initial technical and financial investment, and to produce scientific work around reproducibility, particularly from a methodological standpoint, as illustrated by various publications listed in the R2lab web site.

Worth being mentioned as well, as part of a partnership with the Open Air Interface (OAI) initiative, R2lab is used on a daily basis for system-wide regression tests of the OAI stack, which in return allows us to offer up-to-date images for running OAI-based experiments. Emphasis has been put lately on offering tools that leverage kubernetes as the swiss-knife for orchestrating the deployment of a complete 5G infrastructure as an elastic set of microservices.

In 2021, the management tools on R2lab were extended to support running docker images, in addition to the historic, metal-based image format (.ndz). This allows experimenters to build their images using mainstream tools and base images. Thanks to that, we can now expose the latest OAI code as R2lab-ready docker images, both for (the Evolved Packet Core network) and RAN (the Radio Access Network).

Access to R2lab is open 24/7. R2lab is used by more than 150 users, half of them from France and the other half from all over the world (Australia, Belgium, Brazil, Canada, Chile, Spain Finland, Germany, India, Indonesia, Italy, Japan, Luxembourg, Netherlands Norway, Tunisia, Turkey, UK, US, Vietnam, etc.) to evaluate a wide range of wireless networking scenarios in realistic and reproducible environment. For more details see R2lab home page. This year, two powerful Universal Software Radio Peripheral (USRP) devices were added and are currently available: USRP N300 and USRP N320. Each one is connected through 2x10Gbps SFP+ fibers to our SophiaNode cluster.

Two AW2S 5G Remote Radio Heads (RRH) / Remote Radio Units (RRUs) are also available: a JAGUAR 2T2R RU (CPRI Split 8), IBUmax 50MHz, MIMO 2x2, and a PANTHER 4T4R RU (CPRI Split 8), IBUmax 100MHz, MIMO 4x4. As for USRP N300/N320 devices, each AW2S RU is connected through either 2x10Gbps or 2x25Gbps fibers to our SophiaNode cluster.

The testbed includes also three other 5G modules composed of a Raspberry Pi4 device with a specific hat used to connect a 5G Quectel RM 500Q-GL module using specific kits (composed of M.2/USB3 interface and 4 antennas).

Two new 5G phones replace our old 4G phones: a HUAWEI P40 Pro and a Google Pixel 7. As the previous ones, they are remotely reachable to set up various scenarios.

The future plan is to include R2lab as an addition to the SophiaNode that is developed in collaboration with the OpenAirInterface consortium at Eurecom in the context of the SLICES-RI project.

6.2.2 SophiaNode: an open programmable 5G platform

Our project-team collaborates with Eurecom to deploy and operate an open programmable platform to test post-5G services. Last year, R2lab was connected to a sister site at Eurecom with 600 Gbps fibers forming together the so-called SophiaNode of the ESFRI SLICES-RI project. Last year we enriched R2lab with 5G professional radio units and compute resources managed by Kubernetes clusters to provide an experimental cloud-native environment to test with open source (OAI, SrsLTE) software and some commercially licensed software (e.g. Amarisoft) for 5G/6G networks supporting for example scenarios with disaggregated 5G networks elements. This year we have strongly contributed to making the SophiaNode the reference node for SLICES-RI project. As a result, the SophiaNode is the first node to fully implement the SLICES 5G blueprint. It is used as implementation reference of the other partners. The SLICES blueprint is aimed at sharing the same vision and solutions among the partners as well as proposing a plan for the design and deployment of SLICES RI. In this blueprint we proposed to define the infrastructure baseline augmented with a reference implementation, which aims at keeping a focus on the goals of the project, yet identifying technological challenges and breakthrough at the early stage of the process. Based on developments realized on the SophiaNode, the blueprint shall be deployed by the SLICES-RI partners and will provide the baseline service that will be exposed to the experimenters when SLICES will move into the operation phase (continuous integration and deployment strategy that will start mid 2024).

7.1.1 Leveraging Web browsing performance data for Network monitoring: A data-driven approach

Participants: Chadi Barakat, Naomi Krimi.

Web browsing is one of the most widespread uses of the Internet globally. Despite advancements in network performance, end users still face slow web browsing situations, which can have a range of causes such as a congested Wi-Fi network, a bad wireless signal, or a loaded network and end hosts. It is thus crucial to monitor the network and troubleshoot the specific causes of slow web browsing, as this benefits end users, operators, and internet service providers alike. Various tools attempting to actively collect web measurements through the injection of probes into the network exist. However, no solution is able to identify the specific cause of web browsing performance degradation. This contribution addresses the problem by proposing a new lightweight passive measurement solution capable of transforming web performance measurements collected from within the browser into indicators of network performance and the origin of web browsing anomalies. We validate our solution by emulating a controlled network environment, manually injecting anomalies, leveraging the measurement data available within a browser and building a predictive model that uses a random forest classifier to correctly classify the causes of web browsing performance degradation. Our solution can be used in the wild in the form of a browser plugin to monitor the network and shed light on its anomalies without actively probing it by solely relying on regular user traffic activity. This contribution has been the result of the MIT-Inria internship of Naomi Krimi, and part of the WEMON project that got funded by the Academy of Excellence "Networks, Information, and Digital Society" of Université Côte d'Azur. Within this project a postdoc will join the Diana team in 2024 to work on validating the approach in more realistic scenarios such as cellular networks, and consolidating the machine learning part of the work with further data collection and analysis.

7.1.2 Ray tracing for accurate estimation of signal power and QoS map generation in mobile networks

Participants: Bernard Tamba Sandouno, Chadi Barakat, Walid Dabbous, Thierry Turletti.

Ray Tracing is a propagation modeling approach that accurately estimates the signal power received by end users while considering the details of the environment in their vicinity. The accuracy of this estimation is at the cost of high computational load and high memory consumption due to the heavy computation performed by processes such as the Ray Generation. In this work done in collaboration with the YDATA SME in the context of the PhD thesis of Bernard Tamba Sandouno, we introduce a site-specific ray generation technique able to generate up to 1 million rays within a few seconds and a root mean square error for bandwidth estimation within 2 Mbps. Depending on the location of the antenna, the coverage area, the type of the terrain and the computational resources available, our technique gives the minimum possible number of rays required to accurately estimate end-users' signal power received and their download bitrate. The results of this work were published in 12.

We next build upon this work and use Ray Tracing as a propagation modeling approach for accurate generation of Quality of Service (QoS) maps in mobile cellular networks (signal power, bitrate, etc). The challenge here is that due to the complexity of Ray Tracing, current implementations fail to generate such maps in wide areas. In this second contribution that appeared in 16, always in collaboration with YDATA, we propose an optimization to Ray Tracing able to accurately generate QoS maps in a reasonable time. By leveraging the site-specific ray launching technique outlined above and introducing an innovative alternative to the reception test process tailored specifically for horizontal reception planes, we have achieved a remarkable enhancement in the performance of Ray Tracing. This optimization results in a nearly 1200-fold reduction in execution time, all while maintaining a minimal memory footprint of less than 2%.

7.1.3 YouTube goes 5G: QoE Benchmarking and ML-based Stall Prediction

Participants: Chadi Barakat.

Given the dominance of adaptive video streaming services on the Internet traffic, understanding how YouTube Quality of Experience (QoE) relates to real 4G and 5G Channel Level Metrics (CLM) is of interest to not only the research community but also to Mobile Network Operators (MNOs) and content creators. In this context, and in collaboration with colleagues from University of Campinas in Brazil, we collect YouTube and CLM logs with 1-second granularity spanning a six-month period. We group the traces by their context, i.e., Mobility, Pedestrian, Bus/Railway terminals, and Static Outdoor, and derive key performance footprints of real 4G and 5G video streaming in the wild. We also develop Machine Learning (ML) classifiers to predict objective QoE video stalls by using past patterns from CLM traces. We release all datasets and software artifacts for reproducibility purposes. The study will be published in WCNC 15.

7.1.4 Methodological and ethical recommendations for the evaluation of non-pharmacological interventions: results from a French study using a participatory consensus approach

Participants: Arnaud Legout.

Arnaud Legout explored, in the context of his ElectroSmart project, an experimental therapy based on Open Label Placebo to reduce symptoms of electrosensitive persons. Whereas this project has been abandoned, it introduces Arnaud to the community of Non Pharmacological Intervention (NPI) and to the NPI Society. He joined the expert committee of the NPIS and had a central role in the definition of the statistical methods to evaluate NPIs.

The term non-pharmacological interventions (NPIs) refers to health prevention and care protocols predominantly with a physical, nutritional or psychosocial focus supervised by a professional. Unlike drugs and medical devices, no consensual model for evaluating these complex interventions existed prior to the present study because of large heterogeneity in intervention content and study protocols. This heterogeneity limited the transferability of good practices and led to a great deal of mistrust among professionals and users alike. The present study involved the co-construction of a consensual evaluation framework which meets the specificity of NPIs, participatory approach, and international standards for research in the field of health.

The study involved all French stakeholders, academic and non-academic, researchers, healthcare users, health practitioners, health operators, scientific societies and health authorities. The framework was co-constructed under the direction of a multidisciplinary committee of 22 experts through iterative, open and tracked exchanges, in four successive stages: work performed by a select committee, work performed by a larger committee, open vote by college, and finally, consultation with health authorities and scientific societies.

The framework for evaluating NPIs, called the ‘NPI Model’, includes 14 ethical invariants and 63 methodological invariants shared among five types of study: mechanistic, observational, prototypical, intervention, and implementation. It received the support of 27 learned societies and three French health authorities.

The creation of a standardized framework for evaluating NPIs in the French context has several advantages, as follows: i) it harmonizes and clarifies epistemological, methodological and ethical expectations of a study evaluating NPIs for researchers; ii) it ensures the results of a study are more transferable; iii) it guarantees that programs for professionals in the healthcare, prevention and social assistance sectors are more operational; iv) it facilitates efficient and safer practices for users; v) it provides better understanding for decision-makers and regulation responsible; vi) it ensures more traceable interventions for health operators; vii) it provides solutions that can be better integrated into financing strategies of insurance and social solidarity systems.

Arnaud Legout co-authored this work that is under submission and the official NPI model.

7.2.1 Provable real time network updates

Participants: Damien Saucez.

Next generations of smart factories and industrial systems will rely on commodity Ethernet hardware and 5G. In this context, it is essential to provide means to guarantee that any configuration action in the network preserves network performances within known and acceptable boundaries. By means of binary decision diagrams we are building models of incremental network updates that guarantee not only latency and bandwidth contraints but also jitter and that are immune to vulnerabilities. This ongoing work is done in collaboration with Inria Nancy and Aalborg University.

7.2.2 Exploration and explanation of the Bitcoin address graph structure

Participants: Arnaud Legout.

Bitcoin is the first cryptocurrency blockchain in terms of valuation (+700B USD), it is also the most studied one. However, the internal structure of the graph of addresses is still poorly understood. The goal of this project is to explore several key aspects and this graph.

1. Bitcoin as the first cryptocurrency is used for money laundering and is victim of market manipulation. However, unlike the classical financial markets, all transactions are public. Therefore, we want to identify structural invariants in the graph that represent attempts to either launder money or manipulate the market.
2. The Bitcoin address graph is a large graph with 1 billion addresses and 10 billion edges. However, as there is no preferential attachment, its structure cannot be explained with the same core principles as social networks. We want to explore the structure of this graph and gain new insights into how financial markets start and grow.

This is an on-going work.

7.3.1 SLICES, a scientific instrument for the networking community

Participants: Walid Dabbous, Thierry Parmentelat, Thierry Turletti, Damien Saucez.

A science is defined by a set of encyclopedic knowledge related to facts or phenomena following rules or evidenced by experimentally-driven observations. Computer Science, and in particular computer networks, is a relatively new scientific domain maturing over years and adopting the best practices inherited from more fundamental disciplines. The design of past, present and future networking components and architectures have been assisted, among other methods, by experimentally-driven research and in particular by the deployment of test platforms, usually named as testbeds. However, often experimentally-driven networking research used scattered methodologies, based on ad-hoc, small-sized testbeds, producing hardly repeatable results. We believe that computer networks need to adopt a more structured methodology, supported by appropriate instruments, to produce credible experimental results supporting radical and incremental innovations. In this contribution, we report lessons learned from the design and operation of test platforms for the scientific community dealing with digital infrastructures. We introduce the SLICES initiative as the outcome of several years of evolution of the concept of a networking test platform transformed into a scientific instrument. We address the challenges, requirements and opportunities that our community is facing to manage the full research-life cycle necessary to support a scientific methodology. We participated this year to the SLICES Blueprint workshop in Sorbonne University.

7.3.2 Introducing Fidelity into Network Emulation

Participants: Houssam Elbouanani, Chadi Barakat, Walid Dabbous, Thierry Turletti.

The design and development of new network protocols, architectures, and technologies requires an evaluation phase where the researcher must provide empirical evidence for the performance of their contributions, potentially in comparison to existing solutions. In this context, network emulation has proven to be an attractive approach as it offers more flexibility compared to traditional testing platforms, and more realism compared to simulation. Network emulators provide contained, customizable, and scalable testing environments both for researchers to evaluate their contributions and for the community to reproduce their results. The major drawback of this approach in network experimentation is the use of virtual components (hosts, network switches, etc.) that do not behave with perfect similarity to the physical components they emulate, mainly due to the concurrency in using the underlay network and computing resources. We thus present in this work, that is part of the PhD thesis of Houssam ElBouanani who graduated in March 2023, a methodology to monitor emulation fidelity by measuring the network delays of emulated packets, which relies on statistical metrics to evaluate their inaccuracy. We further dig into the possible sources of emulation inaccuracy and show how our system can detect them to avoid biased experiment results. We particularly show through a common experiment scenario how undetected network emulation errors can lead to biased results. These results were published in 13. In a follow-up work that was published in 14 (and an extented version was published in a journal 11) and always part of the PhD thesis of Houssam ElBouanani 17, we address the problem of troubleshooting distributed network emulation, and propose a heuristic based on linear optimization to extract information about the physical infrastructure from emulation-level packet delay measurements, in order to pinpoint the root causes of emulation inaccuracy with minimal hypotheses. We evaluate the precision of our heuristic using numerical simulations, then show how its implementation performs in a real network scenario.

7.3.3 Adding R2lab worker nodes on the kubernetes SophiaNode cluster

Participants: Damien Saucez, Thierry Turletti.

This demo aims to demonstrate how to use R2lab nodes as workers in our SophiaNode kubernetes cluster. It uses the SLICES blueprint, which automates the kubernetes (k8s) worker node setup using Ansible playbooks . Basically, the demo uses one of the R2lab nodes to launch the ansible playbook that will configure the selected R2lab node(s) for the scenario and attach them to the SophiaNode k8s cluster. All the steps to reproduce the demo, documentation and code are available in GitHub.

7.3.4 Deploying 5G demo on the SophiaNode/R2lab platform

Participants: Thierry Parmentelat, Thierry Turletti.

We have written a nepi-ng script that allows to run 5G OpenAirInterface (OAI) Core Network (CN) and RAN on our new SophiaNode platform. This script aims to demonstrate how to automate an OAI5G deployment on our SophiaNode Kubernetes-based cluster using worker nodes running both on PowerEdge servers on our k8s cluster and FIT R2lab nodes. The script assumes the presence of at minima one FIT node used as k8s worker node attached to the SophiaNode cluster1 that will be used to deploy the OAI-5G scenario on the k8s cluster and within the R2lab anechoic testbed. Different gNB nodes can be used such as USRP B210 connected to k8s worker R2lab nodes, AW2S RRU (such as Jaguar MIMO 2x2 or Panther MIMO 4x4) or USRP N300 or N320 connected with fibers to the oai-gnb pod running on a k8s worker server on the cluster. Regarding UEs, different types are also possible such as 5G Quectel RM 500Q-GL devices attached to either FIT nodes or Raspberry Pi4 nodes in R2lab or 5G phones. In a nutshell, the nepi-ng script first configures the OAI-5G charts to match the parameters of the scenario, e.g., the type of USRP device used to run the gNB and the type of core-network (CN) deployment (basic or advance with different slices). Then it will set up the different nodes used to run the OAI5G functions, including the different UEs selected for the scenario. When all nodes are ready with the required configuration, orchestration of the different OAI5G functions can start through one of the k8s worker R2lab node. All the steps to reproduce the demo, documentation and code are available in GitHub.

7.3.5 Deploying a Multi-access Edge Computing Platform on the R2lab platform

Participants: Thierry Turletti.

We have written a nepi-ng script to deploy the  OpenAirInterface Multi-access Edge Computing Platform blueprint within the R2lab platform. The OpenAirInterface Multi-access Edge Computing Platform blueprint developed by Eurecom was only able to run on simulation mode (using rfsim). We have extended it to make it able to use USRP B210-based gNB and UE nodes (such as Quectel nodes and 5G phones) located in the R2lab platform. In the original blueprint, all the CN, RAN and MEP (Multi-access Edge Computing Platforms) docker containers were running on the same host. This modified blueprint will deploy the core-network, RAN and mep docker compose files on three different FIT nodes. All the steps to reproduce the demo, documentation and code are available in GitHub.

Collaboration with YDATA

We have started a collaboration with the YDATA entreprise on the topic of geolocation assessment of mobile network performance. YDATA is among others the owner of the Zone Adsl website to compare the Internet commercial offers in France. The activity started with a CIFRE thesis. The PhD student Bernard Tamba Sandouno started his PhD on this topic on May 2021. Currently, he is working on developing a set of algorithms to predict the geolocation performance of mobile broadband networks using a ray tracing approach coupled with wireless channel propagation models. The main challenge is in accelerating the rendering of the Radio Frequency performance maps without compromising accuracy and precision as compared to traditional approaches, so that the mapping can scale to the size of a region or a country.

Participants: Chadi Barakat, Walid Dabbous, Damien Saucez, Therry Turletti.

9.1.1 Visits of international scientists

9.1.2 Visits to international teams

Manel Khelifi and Damien Saucez conducted visits to Professor Georg Carle's team at the Technical University of Munich (TUM), who are partners in the SLICES-SC project, on October 24 and November 29, respectively. These visits were specifically focused on the PoS (plain orchestrating service) reproducibility framework, with the objective of seamlessly integrating it into the overall SLICES blueprint. The collaborative efforts during these visits aimed to enhance the synergy between the PoS reproducibility framework and the broader objectives of the SLICES-SC project, contributing to the project's overall success and effectiveness.

Damien Saucez has been invited to the Academic Salon on High-Performance and Low Latency Networks and Systems in Munich on November 30, to present how research infrastructures can be used to enhance reproducibility. In addition a methodology on how to realize provable experiment in real time networks has been presented and discussed with the attendance. A particular focus has been made on how to have exclusive use of real-time hardware resources in shared environments.

9.2.1 Horizon Europe

9.2.2 H2020 projects

9.3.1 PC1, NF MUST : End-to-end multi domain services management architecture of the networks of the future

• Coordinator : Djamal Zeghlache (IMT)
• Inria teams participanting to the project : COATI, DIANA, ERMINE
• Summary : The 5G and 6G end-to-end Multi-Domain Services Management Architecture (NF-MUST) aims at automating production of inter-domain (business and application level) services for 5G, 5G Beyond and 6G networks. A challenging and still unrealized evolution today compared with single domain services or pre-established static multi-domain services that are gradually emerging in 5G and Beyond. Project NF-MUST of the PEPR 5G and Future Networks, focuses mainly on transforming client requests into end-to-end service orderings and on mapping them to resources and network level services (to be) provisioned by the multiple underlying networks. There is a clear evolution of 5 and 6G networks towards the provisioning of services involving multiple players and multiple technologies. Project NF-MUST addresses the related roles and interactions between customers and multiple domains in connection to the other “PEPR 5G and Future Networks” projects, to ensure automated production and operation of multi-domain services across multiple providers. Besides ordering services, NF-MUST will drive the management of the life cycle of the infrastructures” provisioned services and partake in their dynamic and automated adaptation and operation. NF-MUST operates at the business subsystem (BSS) level and at the service side of the operation subsystem (OSS) level. NF-MUST interacts directly with network services treated by project 2 of the overall program.
• Role of the DIANA team: Co-supervise a PhD thesis with COATI on Efficient resource utilisation for service management in next generation networks, to start during 2024.

9.3.2 PC2, NF NAI : Network and infrastructure architectures and network-cloud-sensing convergence

• Coordinator : Gérard Memmi (IMT)
• Inria teams participanting to the project : AGORA, DIANA, RESIST, TRiBE.
• Summary: The capacity to design, develop, plan, and operate networks and convergent network-cloud-sensing systems efficiently, securely and economically in terms of resource consumption. Additionally, these networks and systems will have to be working in collaboration with the various sectors of activity. These networks and systems will also need to be capable of supporting the wide variety of existing applications with heterogeneous resource and performance requirements and sufficient flexibility and agility to dynamically adapt to future needs. The NAI project should go further than traditional objectives like throughput, execution speed, latency, or object connection density and enable the effective integration of a multitude of new technologies. This should include technologies for the physical layer (reconfigurable intelligent surfaces) , 3D transition (NTN - non-terrestrial networks), and architectural principles (like slicing and dynamic end-to-end orchestration). The project will promote the emergence of new applications and services thanks to transparency in terms of performance, robustness, and user security. The project will also put forward and implement interfaces with convergent network-cloud-sensing systems, offering a rich level of transparency to application developers , ranging from the edge to the cloud, from connected mini-objects to large data centres through multi-access edge computing (MEC).
• Role of the DIANA team: Supervise two PhDs to start during 2024. The first one on monitoring plane for mobile cellular networks and the second on experimental evaluation of sliced cellular networks.

9.3.3 PC10, NF FPNG : French network of test platforms for new-generation mobile communications

• Coordinator: Raymond Knopp, Eurecom
• Inria teams participanting to the project : DIANA, Maracas.
• Summary: Setting up national-scale research infrastructures to test new hardware components for 5G and future networks and evaluate the paradigms of next-generation telecommunications networks. The targeted FPNG project is dedicated to setting up nationwide research infrastructures to test new hardware components for 5G and evaluate paradigms for the next generation of telecommunications networks. These research infrastructures will target basic technological components and also end-to-end network testing. This programme of platforms aims to work with all the important technologies in this area - from elementary electronic components to large-scale networking experiments - to provide responses to all the specific challenges defined by the Networks of the Future PEPR project. FPNG's aim is firstly to structure this set of infrastructures and provide free access to existing infrastructures for the national group of PEPR researchers. Its second aim is to invest in new strategic and advanced infrastructures required to respond to the many challenges of the future.
• Role of the DIANA team: participating in the definition and deployment of the infrastructure. Our team has funding designated for acquiring necessary hardware equipment and engaging one DevOps engineer.

10.2.1 Teaching

• Master 2 Ubinet: Chadi Barakat and Walid Dabbous, Evolving Internet, 31.5 hours, M2, Université Côte d'Azur, France.
• Master 2 Ubinet: Chadi Barakat and Walid Dabbous, Internet Measurements and New Architectures, 31.5 hours, M2, Université Côte d'Azur, France.
• Master 1 in Computer Science: Chadi Barakat, Computer Networks, 15 hours, M1, Université Côte d'Azur, France.
• Master 1 in Computer Science: Chadi Barakat, Internet of the future, 15 hours, M1, Université Côte d'Azur, France.
• Master 2 Estel: Chadi Barakat, Voice over IP, 9 hours, Université Côte d'Azur, France.
• Master Ubinet: Arnaud Legout, From BitTorrent to Privacy, 22.5 hours, M2, Université Côte d'Azur, France.
• Master module AWARE (Awarness-Raising to research) : Arnaud Legout, lecture 2 hours, Eurecom, France.
• Thierry Parmentelat helps coordinating the CS courses for the first year students of École des Mines de Paris, covering general topics like Numerical Programming with Python, Advanced Python Programming, and Introduction to Web technologies.

10.2.2 Supervision

10.2.3 Juries

• Chadi Barakat served as reviewer of Ziad Tlaiss's PhD thesis "Automated network packet traces analysis methods for fault recognition and TCP flavor identification", defended in December 2023 at Mines-Telecom Atlantique Bretagne.
• Chadi Barakat served as reviewer of Philippe Graff's PhD thesis "Characterization, in-network identification and optimization of low-latency traffic transport - the case of Cloud-Gaming", defended in December 2023 at Université de Loraine.
• Chadi Barakat served as reviewer of Matthieu Gouel's PhD thesis "Internet-Scale Route Tracing Capture and Analysis", defended in June 2023 at LIP6, Sorbonne Université.
• Chadi Barakat served as jury member for the mid-term review of David Baldassin's PhD thesis "Design and Evaluation of New Transport Protocols for WebRTC", Université Côte d'Azur and Cosmo Softwar, held in May 2023.
• Chadi Barakat served as jury member for the mid-term review of Ilias Driouich's PhD thesis "Privacy-Preserving Algorithms for Federated Learning", Inria, Université Côte d'Azur and Amadeus, held in May 2023.
• Chadi Barakat served as reviewer of Thomas Faval's PhD thesis "Strengthening Privacy and Cybersecurity through Anonymization and Big Data" defended in January 2023 at Politecnico di Torino.
• Chadi Barakat served as reviewer of Hina Qayyum's PhD thesis "Exposing misconduct on free online platforms: A study of mobile apps, web games, and social networks", validated in December 2023 at Macquarie University, Autralia.
• Walid Dabbous served as member of the first year PhD monitoring committe (CSI) for Anderson Lourenço de Araujo at Université Côte d'Azur.
• Arnaud Legout served as jury member for the Ph.D. thesis defense of Rafael Ramos Tubino on 12 december 2023, LIRIS, Lyon 1.
• Thierry Turletti served as reviewer of Mukhtiar Bano's PhD thesis "A Robust Routing Architecture for Hybrid Software Defined and Wireless Mesh Networks", defended in March 2023 at Capital University of Science and Technology, Islamabad, Pakistan.
• Thierry Turletti served as reviewer of Matthews Jose's PhD thesis "In-network real-value computation on programmable switches", defended in March 2023 at Université de Lorraine, LORIA.
• Thierry Turletti served as reviewer of Sagar Arora's PhD thesis "Cloud Native Network Slice Orchestration in 5G and Beyond", in October 2023 at Sorbonne Université en Sciences, EURECOM.

10.3.1 Internal or external Inria responsibilities

DIANA team members actively engage in hosting week-long internships for students at the 3rd level, primarily by introducing them to the state-of-the-art R2lab anechoic chamber. Additionally, they provide insightful explanations of electromagnetic wave transmission in simplified terms, ensuring a comprehensive understanding for the interns.

10.3.2 Interventions

As part of his responsibilities in scientific mediation, Damien Saucez orchestrates visits for the DIANA team researchers, and more broadly, for researchers from the Inria center to high school classes as part of the CHICHE program. Although two scheduled participations by Walid Dabbous and Manel Khalefi were initially planned for 2023, they have been postponed to early 2024 due to scheduling constraints.

International journals

• 11 articleH.Houssam Elbouanani, C.Chadi Barakat, W.Walid Dabbous and T.Thierry Turletti. Troubleshooting Distributed Network Emulation.Annals of Telecommunications - annales des télécommunications2024HALback to text
• 12 articleB.Bernard Tamba Sandouno, Y.Yamen Alsaba, C.Chadi Barakat, W.Walid Dabbous and T.Thierry Turletti. A novel approach for ray tracing optimization in wireless communication.Computer Communications20912023, 309-319HALDOIback to text

International peer-reviewed conferences

• 13 inproceedingsH.Houssam Elbouanani, C.Chadi Barakat, W.Walid Dabbous and T.Thierry Turletti. Delay-based Fidelity Monitoring of Distributed Network Emulation.COMSNETS 2023 :TASIR Workshop - 15th International Conference on COMmunication Systems & NETworkS : Testbeds for Advanced Systems Implementation and Research2023 15th International Conference on COMmunication Systems & NETworkS (COMSNETS)Bangalore, IndiaJanuary 2023HALDOIback to text
• 14 inproceedingsH.Houssam Elbouanani, C.Chadi Barakat, W.Walid Dabbous and T.Thierry Turletti. Troubleshooting Distributed Network Emulation.ICIN 2023 - 26th Conference on Innovation in Clouds, Internet and Networks2023 26th Conference on Innovation in Clouds, Internet and Networks and Workshops (ICIN)Paris, FranceMarch 2023HALDOIback to text
• 15 inproceedingsR. U.Raza Ul Mustafa, C.Chadi Barakat and C. E.Christian Esteve Rothenberg. YouTube goes 5G: QoE Benchmarking and ML-based Stall Prediction.IEEE Wireless Communications and Networking Conference (WCNC)Dubai, United Arab EmiratesApril 2024HALback to text
• 16 inproceedingsB.Bernard Tamba Sandouno, Y.Yamen Alsaba, C.Chadi Barakat, W.Walid Dabbous and T.Thierry Turletti. A Novel Approach to Mobile Outdoor QoS Map Generation.IEEE Conference on Wireless Communications and NetworkingWCNC 2023 - IEEE Wireless Communications and Networking ConferenceGlasgow, Scotland, United KingdomMarch 2023HALDOIback to text

Doctoral dissertations and habilitation theses

• 17 thesisH.Houssam Elbouanani. Introducing fidelity into network emulation.Université Côte d'AzurMarch 2023HALback to text