In this paper an evaluation the detection performances for (AND-CFAR) and (OR-CFAR) processors under different clutter models is done for pulsed radar system.
The clutter models used in this paper are three types of distribution (Exponential Clutter distribution, Rayleigh Clutter distribution and Weibull Clutter distribution).
The two detectors (AND-CFAR and OR-CFAR) are the improved conventional Cell Average-CFAR (CA-CFAR) and Order Statistics CFAR (OS-CFAR) by making full use of the cell information. The two CFAR processors combine the results of the CA-CFAR and OS-CFAR to get a better detection performance.
The mathematical equations of the probability of detection and probability of false alarm to the two detectors (AND-CFAR and OR-CFAR) for the three clutter models are founded under the assumptions that a homogenous background with a Swerling I target, and the reference cells are Independent and Identically Distributed (IID), also the detection performances of these detectors (AND-CFAR and OR-CFAR) have been evaluated for the three clutter models first and compared between them with those of CA-CFAR and OS-CFAR for Rayleigh distribution using MATLAB-Programming (M-File).
In this paper is founded that the detection performance to AND-CFAR and OR-CFAR has not been affected by changing clutter models (clutter probability density function) for fixed probability of false alarm, also for the same probability of detection (Pd=0.7) and for fixed probability of false alarm (Pfa=10-6) and N=12, signal-to-ratio power ration (SNR) which ensure this Pd is equal (18dB) for AND-CFAR, but its equal (18.5dB) for OR-CFAR, this means AND-CFAR is better detection performance than OR-CFAR for different clutter models.