Statistical analysis of measured high resolution land clutter at X-band and clutter simulation

Abstract

A priori knowledge of the clutter statistics is required in maximizing a radar detection range. Much research has been done in analyzing various types of land clutter terrains including urban, forest and crops terrains. However, analysis of palm trees terrain has not been addressed in the open literature. Palm trees have a unique structure, which make them different from other type of trees. Date farms usually consists of date palm trees and scattered small buildings. In this paper, the clutter statistics of date farms terrain were analyzed. The amplitude Probability Distribution Function (PDF) was fitted to different distributions and the Power Spectral Density (PSD) was analyzed using Chan's model. It was found that the amplitude PDF of date farms terrain can be best modelled by a Log-normal distribution. The clutter statistics properties, which can be estimated from measured clutter, can be used in generating simulated clutter samples. Simulated clutter can be used in assessing detection algorithms in the presence of clutter. Generating simulated land clutter with similar statistics to the measured clutter is a challenge. In this paper, simulated clutter samples were generated using a PSD based model. The radar clutter simulator generated complex samples with a PSD similar to the measured clutter, but the amplitude PDF was different. The amplitude PDF of the measured clutter was observed to have a longer tail compared to the simulated clutter.