Clutter-Based Calibration for Multi-Channel SAR Instruments

Kurzfassung

In order to achieve a high resolution and a wide swath, overcoming the contradicting requirements regarding the pulse repetition frequency, near-future satellite Synthetic Aperture Radar (SAR) missions employ multiple channels in azimuth and elevation in combination with Digital Beam-Forming (DBF) techniques. The multiple azimuth channel sampling coherently combines the sub-sampled signals of the azimuth channels. In elevation a narrow receive beam can be steered to the direction of arrival of the signal via Scan-On-Receive (SCORE). SCORE additionally improves the Signal-to-Noise Ratio (SNR) and suppresses range ambiguities. The amplitude and phase behavior of the individual channels is affected by error sources like: imbalances, drifts, coupling, noise, gain compression, changes in the frequency transfer function or ADC errors (aliasing, quantization and jitter). These error sources can lead to inter-channel mismatches, resulting in a degradation of the beamformed pattern, which in turn leads to a worsening of performance, e.g. decreased SNR, degraded resolution, higher range and azimuth ambiguity levels and worsened polarization. Thus, give need for an accurate calibration of the channels. The DBF in elevation requires real-time calibration capabilities, which is not provided by (conventional) state-of-the-art SAR calibration. However, raw data-based methods utilising the radar clutter signal correlation properties between channels are well suited for multi-channel real-time calibration. They are based on the well-known van Cittert-Zernike theorem, stating that the correlation of the field of independent scatterers is proportional to the Fourier transform of the scatterer intensity. Here, (existing) data-based calibration is introduced and it is elaborated on how it can be adapted for DBF SAR by taking spaceborne geometry, transmit signal properties, ambiguities and other effects into account. A refined mathematical model for a multi-channel SAR instrument (focusing on a monostatic planar antenna array SAR) incorporating the latter effects is developed. Additionally, a refined estimation method is provided that exploits all possible correlations between channels. Simulation results showing the calibration-performance are presented.