The general theme of my research deals with the characterization of the joint detection-estimation problem arising frequently both in passive and active systems of measurement (radar, telecoms, sonar, gnss) : the estimation of the parameters of an intermittent source of signal of interest embedded in a permanent noisy environment. This problem can be modelled as a binary hypothesis testing. Under the framework of deterministic parametric modelling, the problem under consideration can be addressed progressively, in terms of theoretical and computational complexity, in two steps :
- the assessment of unconditional estimation performance, that is without a prior detection test, by resorting to lower bounds on estimation performance,
- the assessment of conditional estimation performance, that is including a prior detection test and leading to the characterization of the joint detection-estimation problem.
Extension of these results to a more general framework including random parameters (Bayesian inference) with application to navigation is the topic of my research at ISAE.