"A FPGA-Based Architecture for In-Flight SAR Motion Compensation in UAVs"

Fernando E. Ortiz1, James P. Durbano1, John R. Humphrey1, Petersen F. Curt1, and Dennis W. Prather2

1EM Photonics, Inc.
2University of Delaware


Synthetic Aperture Radar (SAR) methods allow aerial imaging in any weather/lighting conditions using radar beams to illuminate the target zone. This technique is widely used in both satellite and airborne imaging equipment. Recently, SAR capabilities have been added to unmanned aerial vehicles (UAVs), but the erratic path these planes follow can cause blurry images. Fortunately, motion compensation algorithms have been developed to take into account both GPS tracking and previously acquired cartographical information, significantly improving the quality of SAR imagery. However these methods are computationally expensive, forcing this processing to be done on the ground. Furthermore, transmitting the raw radar data down to the ground uses significant power and communication bandwidth, both of which are limited on a UAV. In comparison, if the SAR processing were performed onboard, standard image compression could be used when transmitting data, saving the valuable power and bandwidth resources1. In this paper we propose an architecture for implementing motion compensation algorithms in reconfigurable hardware, allowing the processing of SAR data onboard in real time. This architecture is currently being implemented on a COTS FPGA board2 measuring less than 120 cubic inches, using 25 watts of power and yielding an expected 30X speedup over a 3.0 GHz machine running the same algorithm in software.


  1. http://klabs.org/richcontent/Papers/DSP_Study.pdf
  2. 2 J. P. Durbano, F. E. Ortiz, J. R. Humphrey, P. F. Curt, and D. W. Prather, "FPGA-Based Acceleration of the 3D Finite-Difference Time-Domain Method," in Proceedings of the 11th Annual IEEE Symposium on Field- Programmable Custom Computing Machines. Napa Valley, CA, 2004, pp. 156-163.


2005 MAPLD International Conference Home Page