Résumé
In deterministic parameter estimation, the constrained Cramer-Rao bound (CCRB) is of great interest for analysis and design of a system of measurement in the asymptotic region. However in many problems, the probability density function (p.d.f.) parameterized by the unknown deterministic parameters is expressed as the integral of a joint p.d.f. depending on random variables as well, which prevents from using the CCRB. Based on a framework lately introduced to tackle theses problems, various possible forms of a constrained modified CRB (CMCRB) are introduced and compared with the existing litterature.