Search

Bistatic configurations have awakened a great deal of interest in the recent past among the SAR research community for both airborne and spaceborne sensing. However, bistatic SAR campaigns remain a technical challenge. In this context, current experiments are helping to understand real engineering problems concerning bistatic SAR data acquisitions. In February 2003, DLR and ONERA carried out successfully [1] one of the very first bistatic airborne campaigns. The E-SAR and RAMSES systems were flown in two basic geometrical configurations, both involving parallel flights at the same speed: one defined by along-track acquisitions and the other by across-track acquisitions. These bistatic configurations share with monostatic SAR the invariancy in the along-track dimension, which allows and suggests the using of Doppler domain focussing techniques. This paper puts forward an extension of airborne SAR monostatic focussing algorithms to the azimuth-invariant, bistatic case. A derivation of the point response in range-Doppler domain is presented; the differences between azimuth-invariant, bistatic and monostatic point responses are outlined. Using these results, bistatic focussing solutions based on range-Doppler (RD) [2] and extended chirp scaling (ECS) [3] algorithms are analysed. Particular implementation problems, like Doppler parameter estimation and bistatic motion compensation [4] strategies are also presented. Finally, the algorithms are tested with real data from the DLR-ONERA bistatic airborne campaign.

Here are described the first successes in obtaining bistatic cross-platform interferograms from a two aircraft SAR acquisition campaign performed in February 2003 by the DLR and ONERA using their E-SAR and RAMSES facilities. The main challenging issues in the signal processing involved are the estimation and compensation of phase drift between the two radars master oscillators (bistatic receive signal is demodulated with a local oscillator which is not coherent with the transmitter one) and the determination of the relative aircraft trajectories -i.e. interferometric baseline-accurate to the millimetre (X-band wavelength is 3 cm, hence 1 mm error in relative position shifts the phase by about 10/spl deg/). The high phase quality of the first interferograms obtained is encouraging, but also shows that full endo-clutter moving target indication capability in multistatic environment requires at least two passive SAR receivers in the vicinity of the transmitting active SAR system.