7.1.1 cortexlab-minus

• Functional Description:
  Minus is an experiment control system able to control, the whole lifecycle of a radio experiment in CorteXlab or any other testbed inspired by it. Minus controls and automates the whole experiment process starting from node power cycling, experiment deployment, experiment start and stop, and results collection and transfer. Minus is also capable of managing multiple queues of experiments which are executed simultaneously in the testbed.
• Contact:
  Matthieu Imbert

7.1.2 cortexlab-webapp

• Functional Description:
  The cortexlab web application, which aims at easing platform usage and improving the metadata that we can associate with each experimenter and experiment. This metadata aims at improving the metrics we can gather about the platform's usage The cortexlab web application provides several modules and workflows : - a user management module that allows users to manage their account with a graphical interface. This module also contains two administrator workflows: one to import several user accounts, at the same time, from a json file, which is useful for many use cases, and one to request users to re-validate their accounts, if, for example, the expiration date is outdated, - a booking module: it allows users to book the test bed with a user-friendly graphical interface, instead of the command line. It also allows the user to manage their reservations, - a security module. - a statistics module (developped in 2023) which provides some metrics like the calcul of occupancy and usage ratios on a user selected period.
• Contact:
  Pascal Girard

7.1.3 Cortexlab_LORA_PHY

• Name:
  Multi nodes LORA in GNU radio
• Keywords:
  CorteXlab, GNU Radio, LoRaWAN
• Functional Description:

  Dynamic and customizable LoRa physical layer, derived from the original EPFL LoRa implementation in GNU Radio. More information on this implementation can be found in "Dynamic LoRa PHY layer for MAC experimentation using FIT/CorteXlab testbed", written by Amaury Paris, Leonardo S. Cardoso and Jean-Marie Gorce.

  This adaptation allows end-users to connect any existing upper layer to the physical layer through an easy to use interface using the JSON format, without having to implement the upper layer in GNU Radio.

• URL:
  https://­github.­com/­AmauryPARIS/­LoRa_PHY_Cxlb/
• Contact:
  Leonardo Sampaio

7.1.4 CorteXlab-IoT Framework

• Name:
  Framework for PHY-MAC layers Prototyping in Dense IoT Networks using CorteXlab Testbed
• Functional Description:
  This framework was developed in the project ”Enhanced Physical Layer for Cellular IoT” (EPHYL). It provides a customizable and open source design for IoT networks prototyping in a massive multi-user, synchronized and reproducible environment thanks to the hardware and software capabilities of the testbed. The framework has been extended in 2022, to improve modularity, with the project ADT 3D-SIP.
• Release Contributions:
  Extension to improve synchronization problems, and to support modularity. Independent optimization and development of the PHY layer, the MAC layer and the radio resource management algorithms.
• URL:
  https://­github.­com/­AmauryPARIS/­gr-ephyl
• Publication:
  hal-02150687
• Contact:
  Jean-Marie Gorce

7.2.1 FIT/CorteXlab toward integration in SLICES/Europe

FIT (Future Internet of Things) was a french Equipex (Équipement d'excellence) built to develop an experimental facility, a federated and competitive infrastructure with international visibility and a broad panel of customers. FIT is composed of four main parts: a Network Operations Center (FIT NOC), a set of IoT test-beds (FIT IoT-Lab), a set of wireless test-beds (FIT-Wireless) which includes the FIT/CorteXlab platform managed by Maracas team, and finally a set of Cloud test-beds (FIT-Cloud). In 2014 the construction of the room was done and SDR nodes have been installed in the room: 42 industrial PCs (Aplus Nuvo-3000E/P), 22 NI radio boards (usrp ) and 18 Nutaq boards (PicoSDR, 2x2 and 4X4) can be programmed remotely, from internet now.

As the FIT project development phase ended in 2019 , CorteXlab has seen continued usage as well as further developments. FIT/CorteXlab has been used by both INSA and the European GNU Radio Days (Gnu radio days) for both lectures and tutorials. Several scientific measurements campaigns have taken place in the FIT/CorteXlab experimentation room and are under works at the moment.

Show image description

photography showing the CorteXlab plateform

In 2024, CorteXlab became a part of the SLICES-FR programm funded by PEPR/PC PLATEFORMS, in coordination with Raymond Knopp from Eurecom, and Walid Dabbous from the team DIANA, Inria Sophia. The new deployment is expected by the end of 2025.

Radio Fingerprinting Dataset

• Cyrille Morin , Léonardo Cardoso , Jean-Marie Gorce
• Type: research
• Audience: partners
• Evolution and maintenance: basic
• Free Description: A dataset of recorded RF signals, labeled by the identifier of the radio device that emitted the signal. The signals are raw IQ samples recorded with a USRP in the CorteXlab platform. The dataset contains a set of different transmission scenarios, with various signal types, and emission conditions, none of which contain explicit identifying signals. Its intended use is to train and/or test algorithms for RF fingerprinting, identification of the transmitter of a signal, solely based on the RF signature imposed by the transmitter's electronic components. The dataset is used and described in details in 22.

Wideband Spectrum Sensing Dataset

• Cyrille Morin , Léonardo Cardoso
• Type: research
• Audience: team
• Evolution and maintenance: basic
• Free Description: A dataset of recorded RF signals, labeled by metadata such as time, duration frequency, band, modulation and transmitter for each recieved signal. This was recorded with a USRP inside of CorteXlab, with a mixture of modulation types and transmission pattern. It simulates some real world spectrum usage scenarios such as continuous or bursty transmissions for multiple transmitters, with random frequency allocation. Its intended use is to train and/or test algorithms for wideband spectrum sensing, or localisation in time and frequency of unknown signals, as well as modulation recognition.

Multi-Cell Outdoor Channel State Information Dataset (MOCSID)

• Mohamed el Mehdi Makhlouf , Yamil Vindas Yassine , Maxime Guillaud
• Type: research
• Audience: community
• Evolution and maintenance: lts
• Description: The MOCSID dataset contains synthetic wireless propagation data (a.k.a. channel state information, CSI) designed mainly for the purpose of benchmarking channel charting algorithms such as those developed in the CHASER and INSTINCT European projects. Specifically, CSI time series from several hundreds of realistic pedestrian user trajectories in a multi-cell outdoors (campus) environment have been simulated using the NVIDIA Sionna simulator, based on a 3D scene generated from OpenStreetMap data. The dataset is designed to ensure spatial consistency across the users, and to correctly model overlapping service areas, in order to allow the benchmarking of distributed multi-site channel charting algorithms.

8.1.1 Heterogeneous delay traffic with small packets

This research 34 is related to challenges 1 and 2. To answer the call for a new theoretical framework to simultaneously accommodate random user activity and heterogeneous delay traffic in Internet of Things (IoT) systems we propose coding schemes and information-theoretic converse results for the transmission of heterogeneous delay traffic over interference networks with random user activity and random data arrivals. The heterogeneous traffic is composed of delay-tolerant traffic and delay-sensitive traffic where only the former can benefit from transmitter and receiver cooperation since the latter is subject to stringent decoding delays. The total number of cooperation rounds at transmitter and receiver sides is limited to $D$ rounds. Each transmitter is active with probability $\rho \in [0,1]$. We consider two different models for the arrival of the mixed-delay traffic: in Model 1, each active transmitter sends a delay-tolerant message, and with probability $\rho \in [0,1]$ also transmits an additional delay-sensitive message; in Model 2, each active transmitter sends either a delay-sensitive message with probability ${\rho }_f$ or a delay-tolerant message with probability $1-{\rho }_f$. We derive inner and outer bounds on the fundamental per-user multiplexing gain (MG) region of the symmetric Wyner network as well as inner bounds on the fundamental MG region of the hexagonal model. Our inner and outer bounds are generally very close and coincide in special cases.

8.1.2 When random tensors meet random matrices

This research 35 is related to challenge 1. The proposed study is not directly related to communication problems, but is meaningful for several classes of problems (e.g. massive access, channel charting,...) Relying on random matrix theory (RMT), this paper studies asymmetric order$-d$ spiked tensor models with Gaussian noise. Using the variational definition of the singular vectors and values of (Lim, 2005), we show that the analysis of the considered model boils down to the analysis of an equivalent spiked symmetric block-wise random matrix, that is constructed from contractions of the studied tensor with the singular vectors associated to its best rank-1 approximation. Our approach allows the exact characterization of the almost sure asymptotic singular value and alignments of the corresponding singular vectors with the true spike components. In contrast to other works that rely mostly on tools from statistical physics to study random tensors, our results rely solely on classical RMT tools such as Stein's lemma. Finally, classical RMT results concerning spiked random matrices are recovered as a particular case.

8.1.3 Lossy compression for decision making with risk

Our work in 53 contributes to challenges 1 and 5. Storage or communication of data often requires lossy compression. As this data may be later utilized for decision making, compression introduces a source of uncertainty. In order to verify the quality of decisions, it is often desirable to quantify this uncertainty. This is often achieved via families of risk measures (such as distortion risk measures), rather than the mean. In this work, we study fixed-length lossy compression subject to distortion risk measure constraints. We first establish conditions for the existence of a fixed-length lossy source code, which satisfies a distortion risk measure constraint. We then investigate the impact of quantifying uncertainty via a distortion risk measure, rather than standard expected distortion constraints. Finally, we quantify the impact of changing the source distribution on performance, measured in terms of a distortion risk measure.

8.1.4 Matching and pricing of buyers and sellers in data exchanges

In 54, we consider matching and pricing of buyers and sellers in data exchanges. This topic falls in the context of challenge 5. Buyers are data consumers that wish to obtain data of a sufficient quality to achieve their goals and who may value quality levels differently. Sellers are data providers that desire to properly value the data that they are selling. Our goal in this work is to develop a tractable formulation of the winner determination problem for such a data exchange, and we show that the problem can be solved via bi-level optimization methods. We also examine how different pricing rules are affected when a data provider is able to replicate their data and thus sell it to multiple buyers. We demonstrate that Vickrey and Balanced Winner Contribution rules can introduce inherent disincentives for data replication. Therefore, we introduce a new rule, the modified Balanced Winner Contribution rule, and show that it can provide flexible incentives for data replication in thin markets.

8.1.5 Dynamic stochastic models

In 29, we study stochastic models for biochemical reaction systems with diffusion. This work continues to explore molecular communications as a special case of challenge 5. As described in section 8.2, additional works are done on molecular communications from optimal detection perspective.

One could either use stochastic, microscopic reaction-diffusion master equations or deterministic, macroscopic reaction-diffusion system. The connection between these two models is not only theoretically important but also plays an essential role in applications. This paper considers the macroscopic limits of the chemical reaction-diffusion master equation for first-order chemical reaction systems in highly heterogeneous environments. More precisely, the diffusion coefficients as well as the reaction rates are spatially inhomogeneous and the reaction rates may also be discontinuous. By carefully discretizing these heterogeneities within a reaction-diffusion master equation model, we show that in the limit we recover the macroscopic reaction-diffusion system with inhomogeneous diffusion and reaction rates.

8.2.1 Massive random access optimization with statistical dependencies

Massive access is an important topic for the development of machine to machine communications, in strong connection with 6G and also non terrestrial networks. This is the topic of challenge 2, with strong connections with challenge 1 and challenge 5.

Grant-free random access (GFRA) is a bottleneck to the development of wireless networks supporting the use cases of massive machine-type communication and ultra reliable and low-latency communication. Such networks are densely and massively popupated and must meet stringent latency and reliability requirements. Due to these characteristics, grant-free random access is envisioned to alleviate the control overhead generated by the classical random access procedure.

The originality of our contribution is the introduction of data dependencies (in time and space).

In 28, we extended the work developed in the PhD of Lélio Chetot extending approximate message passing algorithms taking into account stochastic dependencies, focusing on active user detection and channel estimation algorithms. Existing algorithms assume that the activity of each device is homogeneous and independent, which is not the case in many applications (e.g.,due to sensors observing a common phenomenon). In order to address this problem, we introduce a new flexible model taking into account a group-heterogeneous activity, using the framework of copula theory. It is then leveraged by a hybrid generalized approximate message passing algorithm to solve the active user detection and channel estimation problem. Our numerical results show that the user detection and channel estimation are both improved with this new algorithm w.r.t. state-of-the-art Bayesian algorithms, with gains up to 10 times fewer detection errors and 10 dB less channel estimation error. In 32 we worked on designing frame slotted ALOHA algorithms for networks with heterogeneous activity probabilities. Current algorithms for optimizing the probability a device accesses each slot require perfect knowledge of the active devices within each frame. In practice, this assumption is limiting as device identification algorithms in GFRA rarely provide activity estimates with zero errors. In this context, we developped a new algorithm based on stochastic gradient descent for optimizing slot allocation probabilities in the presence of activity estimation errors. Our algorithm exploits importance weighted bias mitigation for stochastic gradient estimates, which is shown to provably converge to a stationary point of the throughput optimization problem. In moderate size systems, our simulations show that the performance of our algorithm depends on the type of error distribution. We study symmetric bit flipping, asymmetric bit flipping and errors resulting from a generalized approximate message passing (GAMP) algorithm. In these scenarios, we observe gains up to $40\%$, $66\%$, and $19\%$, respectively.

In 39 we extend the slot allocation problem above mentioned to a joint power/slot allocation. In 55, specific algorithms have been adapted in the context of 5G. Overall, these contributions are presented in the PhD of Alix Jeannerot 49.

8.2.2 Quantum algorithms for multiple access

This line of research is more explorative, but is also focused on multi-user access (challenge 2). The objective is to explore the use of quantum alogrithms to optimize active user detections (AUD). This work was initiated with the PhD of Idham Habibie, funded by an Inria exploratory action 30. This work has been extended to more realistic scenarios with the PhD of Romain Piron 41, 42, 43, 44. Leveraging quantum annealing (QA) for the AUD problem in massive wireless networks is a promising approach to address the stringent reliability and latency constraints of typical application scenarios. However, the practical implementation of QA on current D-Wave’s processors requires embedding the problem. This increases the number of qubits needed for a given network size, which degrades QA performance. In In 41, we propose to add a pre-processing step called the threshold method to mitigate the undesired effects of embedding. Our results show that, within limited computational time, this threshold method improves QA’s accuracy in solving the activity detection problem. Thus, this is promising to effectively reduce the negative impact of embedding.

42 proposes a new strategy based on Grover’s quantum algorithm to perform the minimum searching so as to implement the ML receiver. As current quantum processors still suffer from noise in the so-called NISQ era, we propose to use Grover’s routine with a reduced number of iterations but with several trials. We show that this approach presents a complexity advantage and allows to reach higher success probabilities than Grover’s. This strategy may permit to timely deploy such AUD receiver even without perfect quantum devices. In complement, in 43, we first propose a mapping between the AUD searching problem and the identification of the ground state of an Ising Hamiltonian. Then, we compare the execution times of our QA approach for several code domain multiple access (CDMA) scenarios. We evaluate the impact of the cross- correlation properties of the chosen codes in a NOMA network for detecting the active user’s set. In 44 we show that the maximum a posteriori decoder of the activity pattern of the network can be seen as the ground state of an Ising Hamiltonian. For N users in a network with perfect channels, we propose a universal control function to schedule the annealing process. Our approach avoids to continuously compute the optimal control function but still ensures high success probability while demanding a lower annealing time than a linear control function. This advantage holds even in the presence of imperfections in the network.

8.2.3 Spike Neural Networks for Wake Up radio

This work has been done in the context of the U-WAKE project, within the PhD of Guillaume Marthe 50, 40. This is one of the core contribution of challenge 4, with contributions in the design of a specific PHY layer (challenge 3).

In the context of the Internet of Things (IoT), one of the greatest challenges lies in energy management. Wake-up Radios (WuR) enable devices to remain in standby mode while consuming minimal energy, activating only upon receiving specific signals. In this work, we propose the use of Spiking Neural Networks (SNNs) as Wake-up Radios (WuR). The neural network's role is to recognize the activation sequence of the targeted node within a bitstream to trigger its wake-up.

Our initial contribution demonstrates the relevance of these networks. Our second contribution involves the investigation and proposal of the Saturating Leaky Integrate and Fire (SLIF) model for WuR design. We proposed leveraging a bio-inspired phenomenon called Synaptic Interaction to create a temporal filter dependent on Inter-Spike Timing (IST). This model's parameters have been analyzed to understand how to adapt its IST ranges. The originality of this contribution lies in introducing a novel method for recognizing temporal sequences in the analog domain.

Subsequently, we explored various SLIF neural network topologies, including linear, diamond-shaped, and multilayer architectures, to understand how networks respond to spike sequences. We established foundational work for future research on neuromorphic networks in low-power IoT devices, particularly in WuRs.

8.2.4 Optimization of Zero energy devices

Complementarily to the former task, and in the same context (low energy IoT applications) of challenge 4, we worked with Orange Labs (project 5G Event Labs and CIFRE) 47, on the design of optimal code sequences tfor zero energy devices (ZED). This work concerns the design of new ultra-low power method for smartphones indoor localization, based on ZEDs beacons instead of active wireless beacons. Each ZED is equipped with a unique identification number coded into a bit-sequence, and its precise position on the map is recorded. An SM inside the building is assumed to have access to the map of ZEDs. The ZED backscatters ambient waves from base stations (BSs) of the cellular network. The SM detects the ZED message in the variations of the received ambient signal from the BS. We accurately simulate the ambient waves from a BS of Orange 4G commercial network, inside an existing large building covered with ZED beacons, thanks to a ray-tracing-based propagation simulation tool. Our first performance evaluation study shows that the proposed localization system enables us to determine in which room a SM is located, in a realistic and challenging propagation scenario.

8.2.5 Systems for molecular communications

We worked on Molecular communications for several years. This research aims at studying communication systems based on molecular properties 31, 37. This work is exploratory and is associated to challenge 1 and challenge 5.

In 31, and in the context of the Internet of Bio-Nano Things (IoBNT), nano-devices are envisioned to perform complex tasks collaboratively, i.e., by communicating with each other. One candidate for the implementation of such devices are engineered cells due to their inherent biocompatibility. However, because each engineered cell has only little computational capabilities, transmitter and receiver (RX) functionalities can afford only limited complexity. In this work, we proposed a simple, yet modular, architecture for a cellular RX that is capable of processing a stream of observed symbols using chemical reaction networks. We proposed two specific detector implementations for the RX. The first detector is based on a machine learning model that is trained offline, i.e., before the cellular RX is deployed. The second detector utilizes pilot symbol-based training and is therefore able to continuously adapt to changing channel conditions online, i.e., after deployment. To coordinate the different chemical processing steps involved in symbol detection, the proposed cellular RX leverages an internal chemical timer. Furthermore, the RX is synchronized with the transmitter via external, i.e., extracellular, signals. The proposed architecture is validated using theoretical analysis and stochastic simulations. The presented results confirm the feasibility of both proposed implementations and reveal that the proposed online learning-based RX is able to perform reliable detection even in initially unknown or slowly changing channels.

In another direction, a challenge for real-time monitoring of biochemical processes, such as cells, is detection of biologically relevant molecules. This is due to the fact that spectroscopy methods for detection may perturb the cellular environment. One approach to overcome this problem is coupled microfluidic-spectroscopy, where a microfluidic output channel is introduced in order to observe biologically relevant molecules. This approach allows for non-passive spectroscopy methods, such as mass spectrometry, to identify the structure of molecules released by the cell. Due to the non-negligible length of the microfluidic channel, when a sequence of stimuli are applied to a cell it is not straightforward to determine which spectroscopy samples correspond to a given stimulus. In 37, we proposed a solution to this problem by taking a molecular communication (MC) perspective on the coupled microfluidic-spectroscopy system. In particular, assignment of samples to a stimulus is viewed as a synchronization problem. Our results show improvements over maximum-likelihood synchronization algorithms in terms of detection performance when there are uncertainties in the composition of the microfluidic channel.

8.2.6 ML based approaches for PHY layer optimization

The use of ML techniques at the PHY layer has been introduced by ten years ago, see e.g. 57. The use of ML in wireless networks is seen as a keystone problem for 6G. The ambition is to optimize transmission techniques, with a reduction of the complexity of channel models, and a reduction of pilots and signaling requirements. In this context, we are collaborating for about 10 years with Nokia Bell Labs on these topics, including the recent PhD of Mateus Mota defended by end of 2023 51 on protocol learning. Another joint PhD is the one of Shashwat Mishra, who published with Noia Bell Labs a contribution on the use of GNN for Cell-free massive MIMO (CFmMIMO) 33. This a promising paradigm to provide uniform coverage in future wireless networks. However, a fully connected CFmMIMO system where all the access points (APs) serve every user equipment (UE) makes it challenging to deploy and scale in real-time due to high computational complexity and increased signaling overhead. In this work, we study the problem of downlink power allocation in partially connected CFmMIMO (p-CFmMIMO) systems using maximal ratio transmission (MRT). We utilize the underlying geometry of the problem to propose a graph representation of the CFmMIMO system and develop a graph neural network (GNN) based power allocation strategy to maximize the minimum SINR in the system. We demonstrate that the proposed GNN model has excellent generalizability to deployment size, radio propagation morphologies, and per-AP serving density. Our GNN can address the power allocation problem in the fully connected case, the partially connected case, and even the cellular case with magnitudes lower computational complexity compared to the conventional numerical solvers. Notably, we show that over a wide range of service scenarios, the model achieves a median spectral efficiency that is within $10\%$ of the optimal second order cone programming (SOCP) solution while requiring 100 times fewer FLOPS.

We also worked on the project 5G Event Labs, on a new technique for joint transmission of messages in broacast channel setting 45. A learning-based approach is introduced to optimize a joint constellation for a multi-user MIMO broadcast channel ($T$ Tx antennas,$K$ users, each with $R$ Rx antennas), with perfect channel knowledge. The aim of the optimizer (MAX-MIN) is to maximize the minimum mutual information between the transmitter and each receiver, under a sum-power constraint. The proposed optimization method do neither impose the transmitter to use superposition coding (SC) or any other linear precoding, nor to use successive interference cancellation (SIC) at the receiver. Instead, the approach designs a joint constellation, optimized such that its projection into the subspace of each receiver $K$, maximizes the minimum mutual information $I(W_k;Y_k)$ between each transmitted binary input $W_k$ and the output signal at the intended receiver $Y_k$. The rates obtained by our method are compared to those achieved with linear precoders.

8.2.7 Channel charting

Channel charting (CC) is an unsupervised learning technique that utilizes channel state information (CSI) to construct a low-dimensional representation of the radio propagation environment through dimensionality reduction, with applications to predictive radio resource management and beam management, proximity detection, context awareness for digital twin applications, geofencing, and improving localization 58. We investigate this topic within the CHIST-ERA CHASER project (see Sec. 10.3.3), and leverage CC for the purpose of sensing in project INSTINCT (see Sec. 10.3.1).

With CSI samples being acquired at high sampling rates (typically around 100 samples per second) by each base station (BS), it is not feasible to store all the generated CSI samples. Furthermore, since it is desirable to apply CC over data collected over long durations (potentially weeks or months), it is reasonable to assume that the total amount of CSI that will be acquired is a priori unbounded. We have developed methods for dynamic, online channel charting 56 which solve the core-set selection problem for the specific case of contrastive learning.

Since in a radio access network scenario, it is typical for one mobile user to be within the coverage are of multiple BS, it is desirable to produce a channel chart representation with is consistent across base stations, we have also obtained results on representation alignment in multi-site CC 46. This method allows to construct a unified channel chart spanning multiple BS, despite the user having completely different CSI when seen from either BS.

8.2.8 Additional contributions

The following references are indicated in our publication list but are not detailed in this report.

• The work in 38, 36 is related to the PhD of Yamil Vindas, done before he joined the team.
• The work in 52 is the scientific report of the PEPR project PERSEUS to which Maracas is contributing.

10.1.1 Inria associate eam not involved in an IIL or an international program

10.1.2 Participation in other International Programs

• Maracas is participating to the COST INTERACT CA20120 (INTERACT) on Intelligence-Enabling Radio Communications for Seamless Inclusive Interactions (2021–2025).
• Leonardo Sampaio , Claire Goursaud are co-supervising Dalil Beknadj , a PhD student from the Univ. of Bejaia, Algeria.

Additionally we have several long-term collaborations including:

• Prof. H Vincent Poor, Princeton University: Long-term collaboration, with recent results on information theoretic limits of multi-user networks obtained in collaboration with H. Nikbakht who had a postdoc in MARACAS followed by another in Princeton.
• Prof. Gareth Peters (UCSB): Long-term collaboration (since 2014) on statistical modelling and methodology, including interference modelling and more recently on data exchange platforms.
• Group of Prof. Robert Schober. (FAU Erlangen): Collaboration centered on molecular communications.
• Dr. Vyacheslav Kungurtsev (CTU Prague): Long-term collaboration (since 2016) on optimization theory and process control
• Dr. Michael Barros (University of Essex, UK): Collaboration on molecular communications.
• Prof. Nir Oren (University of Aberdeen): Long-term collaboration (since 2014) on multi-agent systems, recently on data exchange platforms.

10.2.1 Visits of international scientists

10.2.2 Visits to international teams

10.3.1 Horizon Europe

10.3.2 H2020 projects

10.3.3 Other european programs/initiatives

In the framexork of the CHIST-ERA program :

10.4.1 Inria incentive actions

10.4.2 ANR

10.4.3 BPI France

• BPI – France Relance – 5G Events Labs [Consortium: CEA – Centre de Saclay, Ericsson, Inria, Orange] [2021–24]. The 5G Events Labs project aims to boost the economic activity of the events, culture and sports sectors, around ten major sites in France where Orange and its partners will offer 5G coverage, technological platforms and adapted support enabling companies to leverage these technologies and incubate innovations in the areas of services for attendees and organizers. Maracas contributes in grant-free access solutions for IoT, in collaborative and decentralized estimation algorithms.

10.4.4 Research projects in the framework of Programme Agencies

Malcolm Egan ,Jean-Marie Gorce ,Claire Goursaud ,Maxime Guillaud , Leonardo Sampaio , Matthieu Imbert , Cyrille Morin

Within the national Priority Research Programme and Equipment (PEPR) programme, we are involved in multiple sub-projects of the “Future Networks” PEPR (PEPR-NF), which is affiliated to the program agency “From components to systems and digital infrastructure”, led by CEA. Telecommunication networks represent a key issue for French and European industry, society and digital sovereignty. The French government launched a dedicated national strategy, with the ambition to produce significant public research efforts so the national scientific community contributes fully to making progress that clearly responds to the challenges of 5G and beyond. In this context, the CNRS, the CEA and the Institut Mines-Télécom (IMT) are co-leading the '5G' acceleration PEPR to support upstream research into the development of advanced technologies for 5G+. MARACAS contributes to:

11.1.1 Scientific events: organisation

11.1.2 Scientific events: selection

11.1.3 Journal

11.1.4 Invited talks

• Malcolm Egan gave invited talks at Department Statistics UCSB (2024), HALICIOGLU DATA SCIENCE INST UCSD (2024), Workshop on Distributed Computing in Paris (2024), Interdisciplinary Faculty Development Workshop Bennett University (2024).
• Jean-Marie Gorce gave invited talks at Inria-Japan workshop (July 2024), at NII Tokyo (July 2024), at Korea University (Dec 2024).
• Claire Goursaud gave invited talks at ASILOMAR 2024, at IEEE WF IoT 2024.
• Maxime Guillaud gave invited talks at Linköping university, Banff workshop on Algorithmic Structures for Uncoordinated Communications, Thales SIX, ERICSSON Research France.
• Maxime Guillaud gave two IEEE ComSoc Distinguished Lecture tours in Brazil (with 3 lectures) and Germany (with 2 lectures)
• Maxime Guillaud gave a Tutorial on Channel Charting given at ISWCS 2024

11.1.5 Leadership within the scientific community

• Jean-Marie Gorce is a member of the Scientific Committee of GdR IASIS (CNRS Research Group on Information, Learning, Signal, Image and Vision)
• Maxime Guillaud is a member of the Steering Committee of GdR IASIS (CNRS Research Group on Information, Learning, Signal, Image and Vision)
• Maxime Guillaud is Vice-chair of the EURASIP Signal Processing for Communications and Networking technical area committee, since 2024.

11.1.6 Scientific expertise

• Claire Goursaud was reviewer for the JCJC ANR 2024, expert for the ANR Committee TSIA 2024.
• Malcolm Egan was a Reviewer for the Artificial Intelligence Programme (Bennett University, Delhi) 2024.
• Jean-Marie Gorce is a member of the Scientific Committee of GdR IASIS (CNRS Research Group on Information, Learning, Signal, Image and Vision),
• Maxime Guillaud is a member of the Steering Committee of GdR IASIS (CNRS Research Group on Information, Learning, Signal, Image and Vision),
• Maxime Guillaud is Vice-chair of the EURASIP Signal Processing for Communications and Networking technical area committee, since 2024.

11.1.7 Research administration

• Malcolm Egan was
  • Member council of the CITI Laboratory.
  • Lyon representative for the Inria Mission Jeunes Chercheurs.
  • Member COMI Lyon.
  • Organizer seminar of the CITI Laboratory.
• Jean-Marie Gorce was:
  • Head of Science for the Inria Lyon Centre.
  • Member of the Evaluation Committee of Inria.
  • Member of the board of the EEA doctoral school.
  • In charge of the PEPR-NF preparation, for Inria, member of the PEPR-NF operational committee.
• Claire Goursaud was
  • Deputy director of CITI Lab.
  • CNU 61 member.

11.2.1 Teaching

Maracas members are teaching regularly at the telecommunications department of INSA Lyon. We deliver courses with strong connections with our research activity. The main ones are:

• Bachelor : L Cardoso - Electromagnetism and Wave Physics, 104 eqTD, L2, First Cycle Dept, INSA Lyon, France.
• Bachelor : L Cardoso - Mathematics for Engineering, 60h eqTD, L1, First Cycle Dept, INSA Lyon, France.
• Bachelor : L Cardoso, C Goursaud, J. Hamié - Digital Communications, 80h eqTD, L3, Telecommunications dept, INSA Lyon, France.
• Bachelor : L Cardoso, C Goursaud, Research projetcs - 32h eqTD, L3, Telecommunications dept, INSA Lyon, France.
• Master : JM Gorce, L Chetot - Advanced Digital Communications, M1, Telecommunications dept, INSA Lyon, France.
• Master : JM Gorce, L Chetot - Radio Access Networks, 32h eqTD, M1, Telecommunications dept, INSA Lyon, France.
• Master : C Goursaud - Communications Systems, 32h eqTD, M1, Telecommunications dept, INSA Lyon, France.

11.2.2 Supervision

PhD defended in 2024:

• Alix Jeannerot was supervised by Malcolm Egan and Jean-Marie Gorce
• Guillaume Marthe was supervised bu Claire Goursaud
• Shashwat Mishra was supervised by Jean-Marie Gorce

On-going PhDs :

• Tan Khiem Huynh is supervised by Malcolm Egan and Jean-Marie Gorce
• Andrea Joly is supervised by Leonardo Sampaio and Jean-Marie Gorce
• Mohamed El Mehdi Makhlouf is supervised by Maxime Guillaud and Jean-Marie Gorce
• Claire Mesny is supervised by Claire Goursaud
• Romain Piron is supervised by Claire Goursaud
• Samya Tannir Université Bretagne Sud is cosupervised by Maxime Guillaud and Emmanuel Boutillon
• Shanglin Yang is cosupervised by Jean-Marie Gorce and Guillaume Villemaud

11.2.3 Juries

Participation to PhD committees :

• Maxime Guillaud was on the PhD committees of Khodor Safa (CentraleSupelec), A. Oguz Kislal (Chalmers University of Technology), Taha Yassine (INSA Rennes), and Unnikrishnan Kunnath Ganesan (Linköping University). He was in the mid-term PhD committees of Bastien Barbe (INSA Lyon) and Wissal Benzine (EURECOM).
• Jean-Marie Gorce was on the PhD committees of Adam Mekhiche (U Toulouse), Alice Chillet (U Rennes, Lannion), F. Sagheer (L2S, CentraleSupelec), Jiahui Wei (Insa Rennes), Guodong Sun (Inria Paris)

Participation to recruitment committees:

• Claire Goursaud was member of the CRCN Inria Lyon recruitment committee.
• Jean-Marie Gorce was member of the DR Inria recruitment committee and the CRCN admission committee.

Participation to projects evaluation:

• Claire Goursaud was expert for the ANR Committee TSIA 2024.
• Malcolm Egan was a Reviewer for the Artificial Intelligence Programme (Bennett University, Delhi) 2024.

11.3.1 Participation in Live events

• Cyrille Morin held a stand with a GNU Radio spectrum monitoring demonstration in FOSDEM 2024 (Bruxelles) and Campus du Libre 2024 (Lyon), showcasing both RF spectrum usage and the use of free software in research.

International journals

• 28 articleL.Lélio Chetot, M.Malcolm Egan and J.-M.Jean-Marie Gorce. Hybrid Generalized Approximate Message Passing for Active User Detection and Channel Estimation with Correlated Group-Heterogeneous Activity.IEEE Transactions on Communications2024, 1-1HALDOIback to text
• 29 articleM.Malcolm Egan and B. Q.Bao Quoc Tang. Macroscopic limit for stochastic chemical reactions involving diffusion and spatial heterogeneity.Stochastic Processes and their ApplicationsOctober 2024HALback to text
• 30 articleM. I.Muhammad Idham Habibie, C.Claire Goursaud and J.Jihad Hamie. Quantum Minimum Searching Algorithms for Active User Detection in Wireless IoT Networks.IEEE Internet of Things Journal1112June 2024, 22603-22615HALDOIback to text
• 31 articleB.Bastian Heinlein, L.Lukas Brand, M.Malcolm Egan, M.Maximilian Schäfer, R.Robert Schober and S.Sebastian Lotter. Closing the Implementation Gap in MC: Fully Chemical Synchronization and Detection for Cellular Receivers.IEEE Transactions on Molecular, Biological and Multi-Scale Communications2024. In press. HALback to textback to text
• 32 articleA.Alix Jeannerot, M.Malcolm Egan and J.-M.Jean-Marie Gorce. Exploiting Device Heterogeneity in Grant-Free Random Access: A Data-Driven Approach.IEEE Transactions on Vehicular Technology2024, 1-11HALDOIback to text
• 33 articleS.Shashwat Mishra, L.Lou Salaun, H.Hong Yang and C. S.Chung Shue Chen. Graph Neural Network Aided Power Control in Partially Connected Cell-Free Massive MIMO.IEEE Transactions on Wireless Communications239September 2024, 12412-12423HALDOIback to text
• 34 articleH.Homa Nikbakht, M.Michèle Wigger, S.Shlomo Shamai, J.-M.Jean-Marie Gorce and H. V.H. Vincent Poor. Interference Networks with Random User Activity and Heterogeneous Delay Constraints.IEEE Transactions on Information Theory712December 2024, 1043-1076HALDOIback to text
• 35 articleM. E.Mohamed El Amine Seddik, M.Maxime Guillaud and R.Romain Couillet. When Random Tensors meet Random Matrices.The Annals of Applied Probability341AFebruary 2024HALDOIback to text
• 36 articleY.Yamil Vindas, E.Emmanuel Roux, B. K.Blaise Kévin Guépié, M.Marilys Almar and P.Philippe Delachartre. An asymmetric heuristic for trained ternary quantization based on the statistics of the weights: an application to medical signal classification.Pattern Recognition Letters188November 2024, 37-45In press. HALDOIback to text
• 37 articlex.xuewen qian, S.Stefan Angerbauer, M.Malcolm Egan, M.Marco Di Renzo and W.Werner Haselmayr. A Molecular Communication Perspective on Synchronization of Coupled Microfluidic-Spectroscopy.IEEE Transactions on NanoBioscienceJuly 2024HALback to textback to text

International peer-reviewed conferences

• 38 inproceedingsM.Mathilde Dupouy, Y.Yamil Vindas, M.Marilys Almar, B.Blaise Kévin Guépié and P.Philippe Delachartre. WEAKLY-SUPERVISED SEMANTIC SPACE STRUCTURING: CARDIAC CYCLE POSITION FOR CEREBRAL EMBOLI VISUALIZATION USING CONTRASTIVE LEARNING.2025 IEEE International Symposium on Biomedical Imaging (ISBI)2025 IEEE International Symposium on Biomedical Imaging (ISBI)Houston, United States2025HALback to text
• 39 inproceedingsA.Alix Jeannerot, M.Malcolm Egan and J.-M.Jean-Marie Gorce. Joint Slot and Power Optimization for Grant Free Random Access with Unknown and Heterogeneous Device Activity.EUSIPCO2024 - 32nd European signal processing conferenceEUSIPCO2024 - 32nd European signal processing conferenceLyon, FranceAugust 2024HALback to text
• 40 inproceedingsG.Guillaume Marthe, C.Claire Goursaud and L.Laurent Clavier. Enabling Low-Power Signature Recognition for the IoT with SLIF neurons.EUSIPCO 2024 - 32nd European conference on signal processingLyon, France2024, 1-5HALback to textback to text
• 41 inproceedingsR.Romain Piron, F.Fabian Ganzer and C.Claire Goursaud. Simplified Embedding Scheme for Quantum Annealing Applied to Activity Detection in Massive Wireless Networks.IEEE International Conference on Computer Communications (INFOCOM)London, United KingdomMay 2025HALback to textback to text
• 42 inproceedingsR.Romain Piron and C.Claire Goursaud. Hybrid Grover search for AUD on a NISQ device.EUSIPCO 2024 - 32nd European signal processing conferenceEUSIPCO 2024 - 32nd European signal processing conferenceLyon, France2024, 1-5HALback to textback to text
• 43 inproceedingsR.Romain Piron and C.Claire Goursaud. Quantum Annealing for Active User Detection in NOMA Systems.ACSSC 2024 - 58th Asilomar Conference on Signals, Systems, and ComputersPacific Grove (CA), United States2024, 1-5HALback to textback to text
• 44 inproceedingsR.Romain Piron and C.Claire Goursaud. Scheduling Quantum Annealing for Active User Detection in a NOMA Network.Fifth IEEE International Conference on Quantum Computing and Engineering (QCE 2024), IEEEFifth IEEE International Conference on Quantum Computing and Engineering (QCE 2024)Montréal (Québec), CanadaSeptember 2024HALback to textback to text
• 45 inproceedingsM.Maxime Vaillant, A.Alix Jeannerot and J.-M.Jean-Marie Gorce. Joint Constellation Shaping Using Gradient Descent Approach for MU-MIMO Broadcast Channel.2024 IEEE 25th International Workshop on Signal Processing Advances in Wireless Communications (SPAWC)Lucca, ItalyIEEESeptember 2024, 56-60HALDOIback to text
• 46 inproceedingsY.Yamil Vindas and M.Maxime Guillaud. Multi-Site Wireless Channel Charting Through Latent Space Alignment.International Workshop on Signal Processing Advances in Wireless CommunicationsLucca, ItalySeptember 2024HALback to textback to text
• 47 inproceedingsS.Shanglin Yang, Y.Yohann Benedic, D.-T.Dinh-Thuy Phan-Huy, J.-M.Jean-Marie Gorce and G.Guillaume Villemaud. Indoor Localization of Smartphones Thanks to Zero-Energy-Devices Beacons.2024 18th European Conference on Antennas and Propagation (EuCAP)Glasgow, United KingdomApril 2024HALDOIback to text

Edition (books, proceedings, special issue of a journal)

• 48 proceedingsDistributed Computing and Intelligent Technology - 21st International Conference, ICDCIT 2025.Distributed Computing and Intelligent Technology ICDCIT 202515507Lecture Notes in Computer ScienceSpringer Nature Switzerland2025HALDOI

Doctoral dissertations and habilitation theses

• 49 thesisA.Alix Jeannerot. Uplink Resource Allocation Methods for Next Generation Wireless Networks.INSA lyonDecember 2024HALback to text
• 50 thesisG.Guillaume Marthe. Spiking neurons for wireless communications.INSA de LyonNovember 2024HALback to textback to text
• 51 thesisM. P.Mateus Pontes Mota. Protocol Emergence with Multi-Agent Reinforcement Learning.Université de LyonJanuary 2024HALback to text

Reports & preprints

• 52 reportC.Charbel Abdel Nour, C.Cédric Adjih, K.Karine Amis, X.Xavier Begaud, M.Matthieu Crussière, A.Antoine Durant, M.Marco Di Renzo, C.Catherine Douillard, H.Hajar El Hassani, J.Joumana Farah, I.Inbar Fijalkow, D.Davy Gaillot, J.-M.Jean-Marie Gorce, C.Claire Goursaud, M.M. Guillaud, D.Didier Le Ruyet, M.Mabrouk Asma, P.Pascal Paganini, D.-K. G.Dang-Kièn Germain Pham, B.Balakrishna Prabhu, G.Ghaya Rekaya Ben Othman, E. P.Eric Pierre Simon and R.Rafik Zayani. Deliverable D1 - Technical Report NF-PERSEUS 2023: Power-Efficient Radio interface for Sub-7GHz distributEd massive MIMO infrastrUctureS.CEA - Commissariat à l'énergie atomique et aux énergies alternativesApril 2024, 1-86HALback to text
• 53 miscM.Malcolm Egan. Fixed-Length Lossy Compression with Distortion Risk Measure Constraints.May 2024HALback to textback to text
• 54 miscM.Malcolm Egan and N.Nir Oren. Package Exchange Mechanisms with Quality Constraints for Data Markets.April 2024HALback to text
• 55 miscK. P.K. Pavan Srinath, A.Alix Jeannerot and A. V.Alvaro Valcarce Rial. Joint Resource-Power Allocation and UE Rank Selection in Multi-User MIMO Systems with Linear Transceivers.2024HALDOIback to text
• 56 miscY.Yamil Vindas and M.Maxime Guillaud. Dynamic Channel Charting: Integrating Online Sample Selection with Continual Learning for Streaming CSI Data.November 2024HALback to textback to text