Assessing the Performance of Mean-Level CFAR Detectors in Gamma-Distributed Radar Clutter
Loading...
Date
2023-10-03
Journal Title
Journal ISSN
Volume Title
Publisher
University of M'sila
Abstract
In this study, we addressed the problem of adaptive detection of radar targets in a
homogenous Gamma-distributed clutter. This type of detection is achieved by maintaining a
Constant False Alarm Rate (CFAR) during the detection process. We assume that the radar uses
a square-law device before the CFAR processor window. First of all, we present the
fundamentals of radar detection of targets embedded in noise and giving the basics of CFAR
detection. After that, we discuss in more details some Mean-Level CFAR detectors operating
over a Gamma-distributed clutter, namely the CA- and GO-CFAR detectors. To do so, we
carried out a complete theoretical analysis of both detectors. Indeed, closed-form expressions
for the probability of false alarm (Pfa) have been presented for each detector. The calculation of
the probability of detection (Pd), using the exact statistics of the cell under test (CUT), results in
complicated integrals, which are hard to evaluate numerically. Thus, we propose approximate
expressions for the probability of detection (Pd) for both detectors. These approximations are
easier to compute and straightforwardly implementable in real-time applications. The obtained
theoretical results are then tested and validated numerically by comparing them to their
counterparts computed using numerical integrals and Monte-Carlo simulations, considering
various scenarios. Moreover, a performance comparison of the studied detectors with the
optimal detector has been made assuming a homogenous environment. The obtained results
confirmed the efficacy of the CA-CFAR detector in the case of homogeneous environments
Description
Keywords
Radar, Adaptative CFAR detection, non-Gaussian Clutter, Gamma-Distribution, CA-CFAR, GO-CFAR