This thesis proposal addresses the development of methods for rapid detection and characterization of gravitational events from data collected by the LISA (Laser Interferometer Space Antenna) mission, a space-based gravitational wave observatory scheduled for launch by ESA in 2035. Following the groundbreaking first direct detection of gravitational waves in 2016 and the successful validation of LISA technologies through the LISA Pathfinder mission, the scientific community is preparing for LISA's deployment. The observatory will consist of three satellites separated by 2.5 million kilometers, capable of detecting gravitational waves that are undetectable by ground-based interferometers. A critical component of mission preparation involves developing data analysis pipelines that can rapidly detect and characterize gravitational wave events. 

While rapid analysis methods exist for ground-based interferometers, space-based systems like LISA present unique challenges. The thesis must address: packet-based data transmission requiring event detection from incomplete data, the presence of artifacts such as glitches and non-stationary noise, detection and discrimination of diverse gravitational wave sources, and stringent time and computing constraints.

The thesis focuses on developing a robust and efficient method for early detection of massive black hole binaries (MBHBs). The proposed method will handle the expected LISA data flow, manage potential artifacts, and produce alerts with confidence indices and initial source parameter estimates (coalescence time, binary mass, to only name a few). These rapid estimates are essential for initializing more precise but computationally expensive parameter estimation procedures.

The research will address four key problems: i - implementing a detection method based on sparse representation of MBHB signals through temporal warping, building on promising preliminary results from IRFU; ii - optimizing parameter estimation using machine learning to aggregate information from LISA constellation orbit and measurements; iii - extending the method to accommodate packet-based data transmission; and iv - evaluating the approach on realistic LISA simulations within the Coordination Unit L2A framework, considering scenarios with artifacts and multiple sources.

The work combines statistical methods for data analysis with signal representation through machine learning, incorporating substantial numerical simulation and statistical analysis components. The research also requires understanding different gravitational waveforms and interpreting results in terms of signal detectability within LISA's rapid analysis framework.

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

For more Information about the topics and the co-financial partner (found by the lab!); contact Directeur de thèse - jerome.bobin@cea.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!