RADAR SATELLITE CONSTELLATIONS: SAR CHARACTERIZATION AND ANALYSIS Nathan A. Goodman (1) and James M. Stiles (2) (1) Department of Electrical & Computer Engineering, The University of Arizona 1230 E. Speedway Blvd., Tucson, Arizona 85721 USA, Email: goodman@ece.arizona.edu (2) Radar Systems and Remote Sensing Laboratory, The University of Kansas 2335 Irving Hill Road, Lawrence, Kansas 66045 USA, Email: jstiles@rsl.ku.edu ABSTRACT We discuss the use of radar satellite constellations, or clusters, for wide-area, high-resolution SAR. These systems are composed of multiple, formation-flying satellites with each satellite having its own, coherent receiver. Increased swathwidth compared to that of traditional SAR is attained by processing the data obtained from multiple satellites. The multi-channel system can also be scanned both forward and backward. The size and orientation of a such a system’s resolution cell can change dramatically, however, depending on the number of satellites in the constellation, the size of the constellation, the look geometry, and the subset of data that are coherently processed. In addition, any of these parameters can be varied on demand according to mission requirements. The constellation itself forms an array that is sparsely populated and irregularly spaced. Furthermore, if the constellation is of extremely wide extent, then the width of its array pattern determines resolution rather than system bandwidth and coherent integration length. Forward- and backward-looking scenarios further exacerbate the problem of predicting system resolution. In order to aid in the design, analysis, and signal processing of radar satellite constellations, we present a method of characterizing the resolution of such systems. We derive two eigensensors that can be interpreted as the dimensions of a two-dimensional synthetic aperture. Then, the synthetic aperture expression is used to derive resolution, and simulations are presented to verify the theory. Next, we use results from the synthetic aperture analysis to develop design constraints for this type of radar system. Finally, we use the analysis to suggest areas of future research. 1. INTRODUCTION Synthetic Aperture Radar (SAR) is a powerful, effective tool used for many applications. Radar’s ability to operate in darkness and see through clouds and precipitation has led to its widespread use on both airborne and spaceborne platforms. But unfortunately, the coverage area for a single-aperture SAR is fundamentally limited by the minimum SAR antenna area constraint [1, 2]. Consequently, multiple passes may be required to fully cover a region of interest, which reduces the time available for observing other regions and may lead to time decorrelation of the interferometric phase in multi-pass interferometric SAR (InSAR) [3]. While the ScanSAR technique [1-2, 4] increases SAR coverage through beam scanning, it does so at the expense of azimuth resolution. Fortunately, the minimum SAR antenna area constraint is based on a single-aperture system. If a coherent, multi- channel array is used, then the constraint can still be met while providing increased coverage without sacrificing resolution [5]. This is achieved by using the array pattern to resolve range-Doppler ambiguities. There is currently a strong push toward improving spaceborne radar technology. This is due to several advantages provided by spaceborne radar, the most important of which is the ability to achieve global coverage. However, spaceborne implementations are also problematic, especially concerning the size of the required antennas. The minimum size of a SAR antenna is restricted by the minimum SAR antenna area constraint, and low radial velocities observed by spaceborne platforms require high angular resolution in order to distinguish between moving and stationary targets. Therefore, spaceborne applications require large antennas that can be difficult to deploy and expensive to launch. One proposed concept for space-based radar that decouples the antenna size requirements from the need to launch and deploy a single, monolithic antenna array is to place multiple transmitters and receivers into space, each on their own, small satellite [5-6]. These satellites, called microsats, would fly in a formation called a satellite constellation. Each satellite in the constellation would be able to coherently sample the signal transmitted from each of the transmitters in the constellation. In this way, the constellation would work as a single, virtual radar able to operate in multiple modes including interferometric, SAR, and MTI.