On the probability distribution function for airborne radar clutter

Abstract

Modern airborne radars have stringent clutter-limited performance requirements and have become increasingly more complex and less tractable for mathematical analysis. Moreover, the performance evaluation associated with many detector structures is often intractable by analytical means. So computer simulation is an obvious choice for obtaining quantitative results if the radar is set to operate in hostile scenarios. Radar returns must be statistically modeled to predict the radar performance and optimize the resources. It is well known that as the patch size illuminated in low grazing angles by the modern high-resolution radar decreases, the statistical properties of the resulting backscatter or clutter no longer obey the central limit theorem and tend to deviate from the normal distribution in many practical scenarios. The experimentally measured amplitude probability density function (APDF) of clutter reflections exhibits the larger deviations and is better fitted by APDFs such as lognormal, Weibull, K-distribution, etc., depending on the illuminated surfaces. The work reported here shows the modeling and simulation of diverse clutters and their suitable APDFs for a given surface. It also takes into consideration the variations of scattering coefficient with grazing angle for the given surface. The validity of the simulated data is ensured through standard statistical tests. These statistical characterizations are more helpful in fixing the threshold for optimized performance.