SAR and MTI processing of sparse satellite clusters.

摘要

The concept of radar satellite constellations has been proposed in the literature and is currently under research. These constellations possess the wide angular coverage necessary for performing spaceborne MTI and the spatial sampling necessary for wide-area SAR. However, the constellations form arrays that are sparsely populated, three-dimensional, and irregularly spaced. Therefore, current SAR and MTI processing algorithms, which often assume a single element or a uniform, linear array, are not well suited to this type of implementation.; In order to derive effective SAR and MTI processing techniques, I develop a novel method of characterizing the behavior of sparse-array radar. A radar's five sensor parameters-time, frequency, and 3D spatial location-are equivalently represented by a 2D synthetic aperture. This synthetic aperture is then used to characterize radar properties such as resolution and ambiguity. In addition, I use the synthetic aperture as the basis for a new method of estimating clutter rank when performing MTI. The synthetic aperture is crucial to accurately estimating clutter rank using this technique.; From the synthetic aperture, effective SAR and MTI algorithms are developed. It becomes apparent that traditional SAR matched filtering produces poor results due to the sidelobes of the sparse, irregular array; therefore, other filters are derived to handle the sparse-array case. Simulations are performed for each filtering technique, the performance of the filters versus various system parameters is compared, and impacts of the constellation concept on system design are inferred from the results.; Next, the synthetic aperture is used to estimate clutter rank and define a clutter subspace. MTI is then performed through generalized DPCA filtering, which projects the radar data orthogonal to the predicted clutter subspace. Although DPCA is a mature concept, application of DPCA to a sparse, spaceborne array is new. MTI performance is also analyzed versus system parameters, and conclusions are made concerning the design limitations on sparse, spaceborne MTI systems.