'This thesis presents a number of stochastic models and statistical methods for stochastic channel modeling. The purpose of this is to describe the statistical modeling and estimation of an impulse response function for an ultra wide band radio channel, in a rigorous mathematical context. Previous stochastic models are presented and reformulated using the theory of point processes. The thesis provides brief descriptions of basic likelihood and Bayesian inference. The likelihood section introduces the expectation maximization algorithm, and an approach on how to implement this for inference in the statistical modeling of an impulse response function. The theory of Markov chain Monte Carlo methods and the reversible jump Markov chain Monte Carlo algorithm is presented, for estimation purposes. In the second part of the thesis an algorithm, based on the theory from the first part, is constructed, with the purpose of performing statistical inference on measurements of an impulse response function. The purpose is to fit a parametric model of an impulse response function to measured data.'