Fast and unambiguous direction finding for digital radar intercept receivers.

Abstract

This thesis considers the problem of angle-of-arrival (AOA) estimation in the context of its application to electronic surveillance systems. Due to the operational requirements of such systems, the AOA estimation algorithm must be computationally fast, accurate and will need to be implemented using sparse, large aperture arrays. Interferometry is proposed as a suitable algorithm that meets the operational requirements of electronic surveillance systems. However, for sparse array geometries, phase wrapping effects introduce ambiguities to the phase measurements and so unambiguous AOA estimation requires the use of computationally intensive ambiguity resolution algorithms using three or more antennas. Beamforming and array processing techniques are another class of AOA estimation algorithms that can unambiguously estimate the AOA using sparse, large aperture arrays. While these techniques generally offer better AOA estimation performance than interferometric techniques, they are also comparatively more computationally intensive algorithms. Furthermore, by virtue of using very sparse arrays, high sidelobes in the array beampattern may cause incorrect AOA estimation. This thesis will introduce the concept of using second-order difference array (SODA) geometries which allow unambiguous AOA estimation to be performed in a computationally effcient manner. In the context of interferometry, the so-called “SODA interferometer" will be shown to synthesise the equivalent output of a smaller virtual aperture to allow unambiguous AOA estimation to be performed at the expense of a coarser estimation performance compared to the physical aperture of the array. It will also be shown that the coarse SODA AOA estimate can be used to cue the conventional ambiguity resolution algorithms to provide higher accuracy in a computationally efficient manner. This thesis will also show that the creation of virtual arrays from SODA geometries can be generalised to a larger number of antennas to allow conventional array processing techniques to perform unambiguous AOA estimation in a computationally fast manner. The AOA estimation performance of each algorithm is compared through simulations and also verified using experimental data. This thesis will show that the SODA interferometer, SODA-cued ambiguity resolution algorithms and so-called “second-order array processors" can be used to obtain high accuracy AOA estimates in a more computationally efficient manner than the conventional algorithms.