Clustering and confidence intervals for radar target identification and estimation

Abstract

Pulsed radar data often presents the task of identifying targets and estimating their properties through use of a metric (data-derived, modeled, or hybrid) unique to the targets under study. Accuracy and reliability of this procedure is extremely important, especially since subsequent processing and scientific results can be inaccurate or misleading if target identification and estimation is flawed. For this reason, we present reliable methods for performing these processing steps and quantify their accuracy using real and simulated data.