The first detection of gravitational waves (GWs) from coalescing binary black holes by the
LIGO-VIRGO scientific collaboration has opened a new era in astrophysics: gravitational wave
astronomy. The ground based detectors operate at high frequencies (above 10 Hz), and there are
two major world-wide efforts to detect GW st low frequency. The GWs in the nano-Hz band will
be detected with the Pulsar Timing Array (PTA).
Here we use nature-provided detector: we monitor ultra-stable millisecond pulsars which work
as very accurate clocks. The radio pulses interact with GWs and this interaction is imprinted
as a deviation in the pulses arrival time. In this thesis we will improve existing and develop
new techniques for detecting GWs with PTA. This work could be a part of the first detection of
GWs in the nano-Hz band.

We are developing a space-based observatory (LISA) to detect GWs in mid-low frequencies
(milli-Hz). While LISA will detect quite a different population of GW sources, there are some
detection problems which are common between LISA and PTA. The methods developed for PTA will
also be applied to the simulated LISA data, and, if successful, they will be incorporated in
the LISA data analysis pipeline.

The methods which will be developed during this PhD are based on the Bayesian approach
augmented with machine learning techniques.

This project is supported by ANR grant and the research is conducted in very close collaboration with Orléans-Nançay pulsar group (lead Prof. Gilles THEUREAU). The candidate will also learn data taking pipeline at the radio telescope at Nançay.