Block-Frequency Quantization for Data Volume Reduction in FScan Systems

Abstract:
FScan is a novel acquisition method recently proposed for future SAR systems, which is able to achieve improved resolution and wider swath width compared to conventional SAR systems. This is made possible thanks to the frequency dependant beam pointing of phased array antenna over a large swath in combination with a considerably high transmit bandwidth up to 1 GHz and more. As a result, the target shows a signature in the time-frequency domain which is related to the frequency scan operation. This suggests a convenient data reduction method in which the range spectrum is divided in blocks. In this contribution we propose a complete dedicated onboard processing solution to efficiently encode the recorded data through a novel quantization method named block-frequency block-adaptive quantization (BF-BAQ) as it grounds on the raw data are quantized using state-of-the-art data compression scheme such as BAQ. Simulations have been carried out to deal with the peculiar FScan system characteristics which are presented together with a performance evaluation. For the investigated FScan system scenario, the proposed BF-BAQ achieves a considerable data volume reduction of up to 60% compared to standard quantization methods adopted for spaceborne SAR.