Robust decision making method for adaptive ordered-statistics CFAR technique using information elasticity

Some performance characteristics of ordered-statistics CFAR (OS-CFAR), such as probability of false alarm (PFA) and probability of detection (PD), are controlled by many parameters. Some of these parameters can be considered decision parameters, since they are user defined to achieve a fixed PFA, and an optimized PD. However, other parameters that control these probabilities have the tendency to fluctuate based on the radar environment and operating conditions. These environmental variables can sometimes be difficult to predict and may affect performance. In this paper, a robust decision making method is presented, which selects decision parameters that provide robust performance even in the presence of these variations. The relevant environmental variables investigated in this paper are the number of interfering targets within the detection window and the signal-to-noise ratio (SNR). The Forward Automatic Order Selection Ordered Statistics Detector (FAOSOSD) is used to provide an estimate for the number of interfering targets, and the accuracy of this estimate is observed as a function of SNR. The proposed method defines a performance metric and observes its mean and variance over the uncertain parameter SNR. A trade-off behavior is shown between this mean and variance, and using information elasticity analysis, a decision is selected.