Modeling and simulation of radar sea clutter using K-distribution

Abstract

Modern radars have become increasingly more complex and less tractable for mathematical analysis. So simulation is an obvious prerequisite if the radar is set to operate in hostile scenarios. The very commonly used and well established base band equivalent model for radar clutter is a complex Gaussian process, which implies that Chi-Square for power, or equivalently Rayleigh for amplitude. However, the statistical property of the radar cross-section on the area of sea surface is generally found to be non-Rayleigh. Also that the received clutter results from a large number of independent and identically distributed (IID) elementary scatterers does not hold good. In fact, if only a limited number of such scatterers actually contribute to the received clutter echo, as is the case for high resolution and/or low grazing angles, then the measured amplitude probability density function (APDF) exhibits large deviations from the Rayleigh distribution and is better fitted by families of APDFs such as K-distribution. This paper discusses the modeling and simulation of K-distributed sea clutter to help in understanding the clutter characteristics from a statistical viewpoint. This simulation will lead to a way of optimally fixing the threshold to improve detection performance.