Document Type


Date of Award

Spring 5-31-1998

Degree Name

Doctor of Philosophy in Electrical Engineering - (Ph.D.)


Electrical and Computer Engineering

First Advisor

Alexander Haimovich

Second Advisor

Yeheskel Bar-Ness

Third Advisor

Stanley S. Reisman

Fourth Advisor

Eliza Zoi-Heleni Michalopoulou

Fifth Advisor

Hung Wang


In radar applications, the statistical properties (covariance matrix) of the interference are typically unknown a priori and are estimated from a dataset with limited sample support. Often, the limited sample support leads to numerically ill-conditioned radar detectors. Under such circumstances, classical interference cancellation methods such as sample matrix inversion (SMI) do not perform satisfactorily. In these cases, innovative reduced-rank space-time adaptive processing (STAP) techniques outperform full-rank techniques. The high pulse repetition frequency (HPRF) radar problem is analyzed and it is shown that it is in the class of adaptive radar with limited sample support. Reduced-rank methods are studied for the HPRF radar problem. In particular, the method known as diagonally loaded covariance matrix SMI (L-SMI) is closely investigated. Diagonal loading improves the numerical conditioning of the estimated covariance matrix, and hence, is well suited to be applied in a limited sample support environment. The performance of L-SMI is obtained through a theoretical distribution of the output conditioned signal-to-noise ratio of the space-time array. Reduced-rank techniques are extended to constant false alarm rate (CFAR) detectors based on the generalized likelihood ratio test (GLRT). Two new modified CFAR GLRT detectors are considered and analyzed. The first is a subspace-based GLRT detector where subspace-based transformations are applied to the data prior to detection. A subspace transformation adds statistical stability which tends to improve performance at the expense of an additional SNR loss. The second detector is a modified GLRT detector that incorporates a diagonally loaded covariance matrix. Both detectors show improved performance over the traditional GLRT.



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.