Date of Award
Fall 1994
Document Type
Dissertation - Restricted
Degree Name
Doctor of Philosophy (PhD)
Department
Electrical and Computer Engineering
First Advisor
Richie, James E.
Second Advisor
Heinen, James
Third Advisor
Brown, Donald H.
Abstract
The suppression of interferences is conventionally done by passing the signal that is corrupted by interference signals through a filter (such as an FIR digital filter) which tends to suppress these interferences while leaving the desired signal relatively unchanged. However, if these interferences are very strong compared to the desired signal such as in the case of jamming communications, then conventional filters at the receiver will ultimately fail to remove these strong interferences. In this case the use of the antenna array is essential in eliminating such strong interferences. Thus, the antenna array may be considered as a spatial filter with an antenna pattern of the array equivalent to that of the frequency response of an FIR digital filter. This dissertation describes several new techniques to suppress strong interference signals that usually can not be achieved using conventional filters. In this study, the antenna array at the receiver utilizes the angle of arrival of interference signals together with the geometry of the array elements to make the sidelobe level of the antenna pattern equal to zero (null) in the direction of the interfering signals. The basic idea of the techniques described in this dissertation is to make use of the array geometry to alter the spatial phase differences on the array. This will perturb the original pattern so that nulls can be placed at jammer angles of arrival. In addition the main beam of the original pattern will not be significantly changed. For a linear array, these spatial phases are obtained by perturbing the elevations, positions, and rotations of the array elements. For a circular array, these spatial phases are obtained by perturbing the radial and angular locations of the array elements. The problem of obtaining the new locations of the array elements to place nulls in the antenna pattern is a nonlinear problem with more unknowns (equal to the number of elements) than the number of available equations ( equal to the number of nulls). Analytic solutions exist if a further constraint is imposed on the unknowns. Analysis and computer simulation for Chebyshev arrays with 20 and 36 element have confirmed the ability of these methods to form controllable nulls in the direction of interferences. Furthermore, it is shown that these approaches lead directly to the type of signal processing architecture that is used to move the array to new locations.