Lower Bounds Applied to the Mean-Square Tracking Error of an Amplitude-Comparison Monopulse Radar

摘要

This thesis addresses the problem of estimating the target angle with respect to the boresight of an Amplitude-Comparison Monopulse Radar. The maximum likelihood estimate is used to track the target and produce a mean-square error. This error is approximated from a computer simulated tracking loop and then compared to the Cramer-Rao and Ziv-Zakai bounds. At high signal-to-noise ratios (SNR), the Cramer-Rao bound is useful in lower bounding the mean-square error. At low SNR levels a phenomena known as the threshold effect occurs and the Cramer-Rao bound becomes unreliable as a lower bound. The Ziv-Zakai bound is a tighter bound and can be used to lower bound the mean-square error at very low SNR levels. It also proves useful in determining the SNR level at which the threshold effect occurs. (Author)