Foundations of Semantic Reasoning for Enhanced AI Cooperation in 6G Multi-Agent Communications
6G will integrate 5G and AI to merge physical, cyber and sapience spaces, transforming network interactions, revolutioning AI-driven decision-making and automation and radically changing the overall system’s perception of the foundational concepts of information and reliability. This requires the native-by-design integration of AI and communication system. Current 5G technologies cannot support such change. 5G limits data to be “teleported blindly” along the network without a priori understanding of how informative is for the receiver(s). As a result, AI algorithm outcomes remain limited to sophisticated pattern recognition and statistical correlations. This represent a major limitation of today sense-process-communicate-memorize intelligent information systems.
To support such revolution with AI, the emerging concept of semantic and goal-oriented communications transforms how information is processed by enabling AI to selectively collect, share, and process data based on its relevance, value, or timeliness to the receiver. Unlike 5G’s focus on high-capacity data transport, semantic communications prioritize meaningful, compressed knowledge sharing to enhance AI reasoning, adapt to diverse environments, and surpass current limitations in intelligent decision-making.
This PhD research explores three cutting-edge areas: (1) semantic communication, where today state of the art mostly is focused on AI-driven semantic compression and robustness, (2) integrated communication and sensing, merging data exchange and environmental sensing for resource-efficient applications, and (3) advances in compositional learning and AI reasoning, enabling intelligent systems to process complex, multi-modal data.
This research is focused on the development of abstract concept compositionality models that AI agents can utilize to understand and reason over complex semantic structures. In this context, the PhD candidate will design new methodologies for compositional reasoning that align with the requirements of multi-user, goal-oriented communication. The models will be constructed to enable compositional information exchange where AI agents can intuitively form, exchange, and infer based on compound semantic representations. By focusing on the inherent compositionality and adaptability of semantic exchanges, this research is positioned to support the next generation of intelligent, contextually aware communication systems. These systems will allow for a more precise and meaningful exchange of information between AI agents, enhancing their decision-making and cooperative abilities across a range of applications, from autonomous robotic swarms to networked IoT devices in smart cities and other intelligent environments. The PhD research will benchmark the proposed novel theoretical grounded concepts against current state of the art solutions in semantic communications by numerical simulation.
In-Sensor Computing for MEMS Sensors: Toward an Electromechanical Neural Network
The rise of machine learning models for processing sensor data has led to the development of Edge-AI, which aims to perform these data processing tasks locally, directly at the sensor level. This approach reduces the amount of data transmitted and eases the load on centralized computing centers, providing a solution to decrease the overall energy consumption of systems. In this context, the concept of in-sensor computing has emerged, integrating data acquisition and processing within the sensor itself. By leveraging the physical properties of sensors and alternative computing paradigms, such as reservoir computing and neuromorphic computing, in-sensor computing eliminates the energy-intensive steps of signal conversion and processing.
Applying this concept to MEMS sensors enables the processing of signals such as acceleration, strain, or acoustic signals, with a significant reduction, or even elimination, of traditional electronic components. This has rekindled interest in mechanical computing devices and their integration into MEMS sensors like microphones and accelerometers. Recent research explores innovative MEMS devices integrating recurrent neural networks or reservoir computing, showing promising potential for energy efficiency. However, these advancements are still limited to proof-of-concept demonstrations for simple classification tasks with a very low number of neurons.
Building on our expertise in MEMS-based computing, this doctoral work aims to push these concepts further by developing a MEMS device that integrates a reprogrammable neural network with learning capabilities. The objective is to design an intelligent sensor that combines detection and preprocessing on a single chip, optimized to operate with extremely low energy consumption, in the femtoJoule range per activation. This thesis will focus on the design, fabrication, and validation of this new device, targeting low-frequency signal processing applications in high-temperature environments, paving the way for a new generation of intelligent and autonomous sensors.
On-line monitoring of bioproduction processes using 3D holographic imaging
The culture of adherent cells on microcarriers (MCs) is a promising approach for various bioproduction applications, such as drug manufacturing and delivery, regenerative medicine, and tracking of cellular differentiation. However, the analysis of single cell morphology and behavior without affecting the substrate integrity remains a major challenge. Lens-free holographic imaging is emerging as a promising solution for real-time, non-invasive monitoring of cellular processes. This technique captures wide field of view images without requiring exogenous labeling or sample manipulation, thus preserving the integrity of the cellular environment.
This thesis proposes the development of a 3D lens-free imaging system to monitor cells on MCs in near real-time. The microscope will be coupled with advanced algorithms for data reconstruction and analysis in on-line bioreactors. The use of deep learning techniques will allow for real-time segmentation and analysis of single cells, facilitating the tracking of cellular dynamics. This innovative project paves the way to a non-invasive monitoring of 3D multicellular samples, with potential applications on organ-on-chip and more complex organoids systems.
Plasma Etching development for the advanced nodes using SADP techniques
The miniaturization of the electronics components involves the development of new processes. Indeed, the 193nm immersion lithography alone does not permit anymore to achieve the dimensional requirements of the most advanced technological nodes (=10nm). Since the last 10 years, multi-patterning techniques have been developed to overcome the i193nm lithography limitations. Herein, we will study the « Self-Aligned Double Patterning » (SADP) technique that divides by two the initial pitch of the lithographical patterns. This technology relies on a conformal deposition of a dielectric film (spacer) over the initial patterns (mandrel). The spacers will be then used as a mask during the pattern transfer by plasma etching. The small targeted dimensions require a perfect control of the etching processes. However, the etching steps can damage the materials used herein leading to a dimension loss. One of the main challenge will be to control the etching steps and so the plasma-induced modification in order to satisfy the specifications (dimension, profile, material consumption, etch rate, uniformity…). Besides, the goal will be also to propose new SADP approaches allowing us to generate different type of patterns in order to produce planar FDSOI transistors, which is currently little reported in literature.
The challenges of this PhD ?
To develop innovative etching processes
To explore new couple of material (spacer/mandrel) and to propose an industrial integration flow that will be validated by electrical tests
To identify the technological obstacles and to propose solutions for overcoming them
To put in place a reliable characterization protocol in order to detect the physical and chemical modifications of the materials used and to accurately measure the final patterns’ dimensions
Sperm 3D - Male infertility diagnostic tool using holography for imaging and 3D tracking
Infertility is a growing problem in all developed countries. The standard methods for the diagnostic of male infertility examine the concentration, motility and morphological anomalies of individual sperm cells. However, one in five male infertility cases remain unexplained with the standard diagnostic tools.
In this thesis, we will explore the possibility to determine the male infertility causes from the detailed analysis of 3D trajectories and morphology of sperms swimming freely in the environment mimicking the conditions in the female reproductive tract. For this challenging task, we will develop a dedicated microscope based on holography for fast imaging and tracking of individual sperm cells. Along with classical numerical methods, we will use up-to date artificial intelligence algorithms for improving the imaging quality as well as for analysis of multi-dimensional data.
Throughout the project we will closely collaborate with medical research institute (CHU/IAB) specialized in Assisted Reproductive Technologies (ART). We will be examining real patient samples in order to develop a new tool for male infertility diagnosis.
Impact of plasma activation on reliability of Cu/SiO2 hybrid bonding integrations
In recent years, CEA-LETI emerged as a leading force in the development of advanced microelectronic manufacturing processes. A key focus has been on wafer-to-wafer Cu/SiO2 hybrid bonding (HB) process, an emerging technology increasingly employed for producing compact, high performance and multifunctional devices. Before bonding, a crucial surface activation step is necessary to enhance the mechanical strength of the assembled structures. Different approaches have been developed, and the most used in the industry is N2-plasma activation. However, this process remains controversial due to undesirable effects, the formation of Cu nodules at the bonding interface between particularly electrical pads and the passivation of Cu pads with chemical complexes. These issues can significantly compromise the electric properties and reliability of devices. In collaboration with STMicroelectronics and IM2NP, this PhD aims at studying the impact of plasma activation on Cu/SiO2 HB integrations.
Ultra-wide-field smart microscope for the detection of egg parasite (SCREENER)
In most parasitic cycles, the free phase passes through an egg stage, which is released by the host into the environment via a complex faecal matrix, which has highly variable and often low egg concentrations. The classical detection method relies on microscopic observation of these eggs, which implies a tedious and time-consuming preparation of the sample to concentrate the eggs, with highly variable sensitivity values. This detection is crucial because, once dispersed, the eggs contaminate the environment and food, leading to cases of parasitic zoonoses in humans.
Detection in environmental and food matrices is even more complex than for faeces because of the very low number of eggs present : 1 to 10 per sample in the vast majority of cases. The thesis aims at developing a lensless wide-field imaging system that will allow the counting and identification of parasite eggs in complex matrices, while increasing sensitivity. This will make it possible to automate detection, thus opening up the possibility of investigating more samples for better health surveillance.
Fracture dynamics in crystalline layer transfer technology
Smart Cut™ is a technology discovered at CEA and now industrially used for the manufacture of advanced substrates for electronics. However, the physical phenomena involved are still the focus of numerous studies at CEA. In Smart Cut™, a thin material layer is transferred from one wafer to another using a key fracture annealing step upon which a macroscopic fracture initiate & propagates at several km/s [i].
____________
Improving technology requires a solid understanding of the physical phenomena involved in the fracture step. The aim of this PhD project is thus to address the mechanisms involved in fracture initiation, propagation and post-fracture vibrations
____________
On the CEA-Grenoble site, with industrial interest, the student will use and further develop existing experimental setups to investigate the fracture behavior in brittle materials, including optical laser reflections [iv], time-resolved synchrotron diffracting imaging [iii], and ultra-fast direct imaging [ii].
In addition, python-based data analysis algorithms will be developed to extract quantitative information from the different datasets. This will enable the student to determine involved mechanisms and evaluate the influence of the wafer processing parameters on the fracture behavior, and thus propose improvement methods.
References :
[i] https://pubs.aip.org/aip/apl/article/107/9/092102/594044
[ii] https://journals.aps.org/prapplied/abstract/10.1103/PhysRevApplied.15.024068
[ii] https://journals.iucr.org/j/issues/2022/04/00/vb5040/index.html
[iv] https://pubs.aip.org/aip/jap/article/129/18/185103/158396
Microwave Near Field Sensing in Heterogeneous Media
This thesis focuses on the development of microwave near-field sensing techniques for applications in biomedicine, agronomy, and geophysics. The primary objective is to design low-complexity algorithms that effectively solve complex inverse problems related to the characterization and detection of dielectric properties with various geometric distributions in heterogeneous media.
The candidate will begin by conducting a comprehensive review of existing radar-based and advanced signal processing methods. A precise physical model of microwave propagation in near-field conditions will be developed, serving as the foundation for new detection methods based on the concept of physics-driven iterative tomography. The ultimate goal is to formulate efficient algorithms suitable for real-time applications and validate them through experimental implementation. To achieve this, an evolving prototype setup will be developed, progressing from 2D media to more complex 3D scenarios.
This interdisciplinary project combines physical modeling, algorithm development, and practical experimentation. It presents an opportunity to advance the field of microwave imaging, with significant implications for biomedical and environmental applications.
High-isolation power supply
With the rapid evolution of technologies and the growing challenges of miniaturization and resource management, power converters are facing ever more stringent performance requirements. To meet these needs, the use of wide-bandgap semiconductors such as SiC (silicon carbide) and GaN (gallium nitride) is becoming increasingly common. These materials significantly increase the switching speed of converters, reducing losses and improving efficiency.
However, this switching speed brings additional challenges: the steepness of the switching edges can cause stray currents that interfere with switch controls. To counter these undesirable effects, it is necessary to use switch drivers offering a high level of insulation. The traditional solution is based on high-frequency magnetic transformers, but these devices are expensive, take up a lot of space and offer limited insulation.
Thesis objective: the aim of this thesis is to design a new solution for powering wide-gap component drivers, by replacing magnetic transformers with piezoelectric transformers. This innovative approach aims to reduce costs, space requirements and improve the overall efficiency of power conversion systems.
Supervision and ressources: the selected candidate will work as part of a leading-edge research team, renowned for its expertise in the field of power conversion using piezoelectric resonators. The team has the resources and know-how to support the development and validation of this innovative technology.