Real-Time Imaging Processor for Highly Squinted Airborne SAR Based on Multicore DSPs

In consideration of highly squinted well-focused synthetic aperture radar (SAR) imagery at a faster speed, a high-performance SAR real-time processing system based on multicore digital signal processor (DSP) is established in this paper. First, the polar format algorithm (PFA) using the principle of chirp scaling (PCS) is implemented for higher algorithm efficiency. Then, autofocus technique is employed for compensating the azimuth phase error. Next, architecture of parallel processing between two DSPs is designed, which not only reduces the time consumption, but also makes enormous SAR data processing possible. Finally, four DSPs in pairs are applied to achieve Ping-Pong nested pipeline processing for meeting the requirements of real-time imaging. Experimental results utilizing the measured highly squinted airborne SAR data are performed to validate the formulated DSP system.