Based on Paris-Saclay University campus, gathering people from L2S, IAS, SATIE, and CEA.

The working group focuses on computational imaging for radio astronomy in relation to the Square Kilometre Array (SKA) project, the largest radio interferometer set to deploy its first phase in 2024. This project, which involves high-performance computing to process 7.2 Tb/s of raw data, brings together experts from Paris-Saclay across various disciplines. Participants include astrophysicists and researchers in digital sciences, data science, and artificial intelligence. Several laboratories are involved, such as L2S, SATIE, CEA, and IAS, contributing expertise in statistical signal processing, algorithm optimization, data calibration, and high-performance computing. The group aims to foster collaboration through regular seminars and a funded Master’s internship to explore advanced computational imaging and statistical methods.

The Working Group is supported by the Graduate School ISN of Paris-Saclay University.

L2S

Signals and Systems Laboratory has recognised expertise in statistical signal processing, antenna processing, inverse problems and computational imaging. The laboratory has worked on the development of new statistical signal processing methods, Bayesian and variational approaches, robust statistics and statistical learning for image estimation and reconstruction, particularly in inverse problems. The laboratory is involved in the SKA project through the ANR Dark-Era (2021-2025) project, which focuses on HPC simulation for the design of a heterogeneouscomputing cluster for SKA, as well as several theses on data calibration and the acceleration of image reconstruction algorithms and the ECLAT (Extreme Computing Laboratory for Astronomical Telescopes) joint laboratory

SATIE

SATIE draws on its expertise in both signal processing and algorithm-architecture matching to meet the computational challenges of new-generation radio telescopes . In signal processing, the work focuses on imaging techniques, dynamic imaging, calibration, the search for optimal geometries, and the switching of sub-networks and informed learning. Work on matching algorithms to architecture aims to better understand and scale up algorithms. Working closely with L2S, SATIE brings this dual expertise to LabCom ECLAT and the ANR Dark-Era consortium. SATIE is committed to the application side of its work, and is developing a collaboration with the Nançay Observatory (ORN), where NenuFAR, an SKA precursor, is located.

CEA/Dedip & CEA/Dap

Three Irfu departments at the CEA are or will be involved in SKA: the DPhP (Department of Particle Physics), the DAp (Department of Astrophysics) and the DEDIP (Department of Electronics, Detectors and Computing for Physics). Radio observations are of interest to several Irfu teams studying cosmic ray sources and high-energyphotons, cosmology, HI (neutral atomic hydrogen) readings at 21cm and the study of compact objects. DEDIP and DAp are also involved in developing algorithmic methods for processing visibilities and reconstructing sky maps, as well as designing the architecture of the SKA Regional Centres and adapting/optimising the use of High-Performance Computing (HPC) resources in this context.

IAS

The IAS cosmology team is participating in the preparation of the SKA via its Cosmic Magnetism and Epoch of Reionisation scientific working groups and via the exploitation of its precursors (notably HERA). The researchers involved in this WG are developing methods for analysing SKA radio data based on Fourier analysis, component separation and inverse problem solving using Bayesian and machine learning methods. These methods have been developed not only to extract physical information from the data, but also to facilitate calibration and the processing of polluting signals such as radio interference. This involves, for example, reconstructing missing information from sparse data or extracting the topology of such maps. The IAS is also involved in the design of SKAO’s French data centre (SRC), as well as institutionally via the Observatoire des Sciences de l’Univers de l’Université Paris-Saclay and the scientific council of the SKA-LOFAR Specific Action.