Radar coverage prediction over ocean: Duct mapping using least squares support vector machines
Résumé
Sea environment is a complex and changing medium which causes important variations on shipborne radar coverage. Actually, the electromagnetic wave propagation sustains strong refractive effects due to the presence of ducts, and diffraction effects due to the sea surface. However, efficient two-dimensional prediction models for wave propagation based on parabolic wave equation (PWE) are available nowadays, but a relevant description of the refractive index profile is needed to feed these models. The idea, first introduced by Rogers and Gerstoft and known as "refractivity from clutter" (RFC), is to exploit the sea clutter in order to retrieve the refractive index profile. The aim of this paper is to propose a new inversion method to obtain the vertical variations of the modified refractivity from the knowledge of the range-dependent propagation losses. These latter can be deduced from the sea clutter collected by the radar. Whereas Rogers and Gerstoft used a genetic algorithm (GA) to find the modified refractivity profile parameters, least squares support vector machines (LS-SVM) are used here. This method produces good results with the great advantage of quickness
Mots clés
electromagnetic wave propagation
least squares approximations
marine radar
parabolic equations
radar clutter
radar computing
wave equations
duct mapping
least squares support vector machine
ocean
parabolic wave equation
radar coverage prediction
refractive effect
refractivity from clutter
sea clutter
sea environment
sea surface
shipborne radar coverage
Clutter
Ducts
Electromagnetic propagation
Electromagnetic refraction
Least squares methods
Oceans
Predictive models
Radar
Refractive index
Support vector machines