Extended title: Analysis and design of periodic structures using AI. Application to polarizing screens and dichroics.

Context: 

Satellites in geostationary orbit have long been the major pillar of the Satcom industry, supporting end users around the world with voice, data, and video services. Breakthrough ‘new space’ services arising from commercialisation of large-scale constellations positioned closer to the Earth in non-geostationary orbit intend to disrupt through low-latency services and pole-to-pole coverage, like Starlink or OneWeb. In this highly-competitive area, developing flat panel user-terminal antennas at K/Ka band represents a major challenge for mass-market deployment. On the other hand, for more than 10 years, IETR has been developing innovative broadband flat antennas, based on Continuous Transverse Stub (CTS) architectures. The latter share a unique radiating aperture in transmit and receive modes, but operate in single-linear polarization. The development of dual-band (K/Ka) and wide field-of-view (up to 60 deg.) polarization converters based on periodic metasurfaces with switchable circular polarization still constitutes a major challenge. 

In addition, observation satellites provide scientists with the essential data needed to detect environmental changes on Earth. As more than 50% of climate variables can only be measured from space, Earth observation satellites are a vital tool for monitoring the effects of climate change on natural ecosystems. In this context, microwave/millimeter wave radiometry is a key tool for monitoring and understanding the environment and climate change, with major challenges in the design of ultra-low loss multi-band dichroics. 

Finally, Artificial Intelligence (AI) has recently emerged as a transformative tool in computational electromagnetics (EM), offering powerful data-driven strategies to complement traditional physics-based solvers. By learning the relationships between structural parameters and EM responses, AI models can drastically reduce simulation time and computational cost, enabling rapid exploration and optimization of large design spaces. This can be especially advantageous for analyizing complex EM structures with many degrees of freedom [1]. For this reason, the synergy between AI and EM is crucial for modern space communications and Earth observation relying on advanced EM structures and materials.

[1] A. Massa, D. Marcantonio, X. Chen, M. Li, and M. Salucci, “DNNs as applied to electromagnetics, antennas, and propagation - A review,” IEEE Antennas Wireless Propag. Lett., vol. 18, no. 11, pp. 2225-2229, Nov. 2019 (DOI: 10.1109/LAWP.2019.2916369).

- Objectives: 

The objective of this thesis project is to exploit innovative AI methods to design two main types of periodic structures: 1) broad-band/dual-band linear-to-circular polarization screens with wide field-of-view (up to 60 deg.) and switchable circular polarization at K/Ka band; 2) Dichroics at millimeter and sub-millimeter waves. Both use cases are extensions of joint activities between IETR and CNES [2],[3].

[2] Projet R&T CNES avec l’IETR, référencé « R-S22/TC-0006-019 – Antennes pour terminal utilisateur Ka faible cout a fort dépointage angulaire », et clôturé le 31/08/2025.

[3] Etude métier CNES-IETR sur les grilles à séparation de polarisation.

- Description of the research program: 

The proposed work program is organized into 5 main stages:

• Stage 1. State-of-the-art on (1) linear-to-circular polarization screens based on periodic structures, (2) dichroics for Earth observation, and (3) AI-based methods. Duration: 2 months.

• Stage 2. Development of a AI-based design framework for the design of periodic structures illuminated by a plane wave. Validation using already-available structures at IETR (e.g. single-band (Ka/Tx) linear-to-circular polarizer, see [2]) Duration: 10 months.

• Stage 3. Design of a dual-band (Ka/Tx) linear-to-circular polarizer using the AI-based methods developed at stage 2. Duration: 12 months.

• Stage 4. Design of one dichroic for microwave/millimeter wave radiometry. The specifications will be inspired by [3]. Duration: 8 months.

• Stage 5 (partly in parallel to stages 3 and 4). Prototyping and experimental characterization of the polarizer and dichroic designed at stages 3 and 4 respectively. Duration: 4 months.

The thesis manuscript will be written in parallel to stage 5.

=================

For more Information about the topics and the co-financial partner (found by the lab!); contact Directeur de thèse - ronan.sauleau@univ-rennes.fr

Then, prepare a resume, a recent transcript and a reference letter from your M2 supervisor/ engineering school director and you will be ready to apply online before March 13th, 2026 Midnight Paris time!