Clutter characterization based on random walks

Abstract

The compound-Gaussian processes have been successfully used in the literature for modelling the sea-clutter returns. We propose a new interpretation of the compound model as a limiting form of a discrete-time random walk with a stochastic number of steps. Since the non-Gaussian properties of the overall process are ruled by the statistics of the appearance and disappearance of the scatterers within the radar cell only, accurate models for such fluctuations are also necessary. The generalized birth and death model leads to a wide class of non-Gaussian-Markov processes which are mathematically treatable and easy to generate by pseudo-random procedures.