Résumé : We propose a new linear-regression model for the estimation of the path-loss exponent and the parameters of the shadowing from the propagation-loss data collected by the mobiles with respect to their serving base stations. The difficulty consists in deriving the parameters of the distribution of the propagation loss with respect to an arbitrary base station from these regarding the strongest one. The proposed solution is based on a simple, explicit relation between the two distributions in the case of infinite Poisson network and on the convergence of an arbitrary regular (in particular hexagonal) network to the Poisson one with increasing variance of the shadowing. The new approach complements existing methods, in particular the one based on COSTWalfisch-Ikegami model, which does not allow for the shadowing estimation and is not suited for indoor scenario.