Performance and processing of SAR satellite clusters

Spaceborne radar systems have been proposed that use the concept of "formation flying" satellites; a system wherein a cluster of small, individual satellites orbit the Earth while remaining in the same relative positions with respect to each other. Each individual satellite would be a standard SAR sensor, but the data collected from each could be combined to produce a single "virtual" satellite from the overall array of sensors. The question from a radar standpoint is, how should this data be processed, and what are the benefits in radar performance resulting from this design? A spaceborne array of radar satellites has the advantage that angle-of-arrival information, in addition to range-Doppler data, is collected. Essentially, the array of satellites can be viewed as a large, albeit sparse, array of antennas. This paper demonstrates that this extra information can be used to produce SAR imagery over both an arbitrarily wide swathwidth and with fine resolution. This is accomplished by applying a minimum mean squared error processing algorithm to combine the data from each sensor. The results of several simulations are shown; they demonstrate the efficacy of the concept and the processing. It is also shown that, as the number of satellites increases, the performance of the MMSE processing increases and the hardware requirements associated with each radar are eased.