Venter, CJGrobler, HAlMalki, KA2012-02-282012-02-282011-12Venter, CJ, Grobler, H and AlMalki, KA. Implementation of the CA-CFAR algorithm for pulsed-doppler radar on a GPU architecture. IEEE Jordan Conference on Applied Electrical Engineering and Computing Technologies (AEECT), Faculty of Engineering and Technology, University of Jordan, Amman, Jordan 6-8 December 2011978-1-4577-1083-4http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?arnumber=6132514http://hdl.handle.net/10204/5609© 2011 EEE. This is the accepted version of the work. Reprinted, with permission, from Venter, CJ, Grobler, H and AlMalki, KA. Implementation of the CA-CFAR algorithm for pulsed-doppler radar on a GPU architecture. IEEE Jordan Conference on Applied Electrical Engineering and Computing Technologies (AEECT), Faculty of Engineering and Technology, University of Jordan, Amman, Jordan 6-8 December 2011. This material is posted here with permission of the IEEE. Such permission of the IEEE does not in any way imply IEEE endorsement of any of CSIR Information Services' products or services. Internal or personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution must be obtained from the IEEE by writing to pubs-permissions@ieee.org. By choosing to view this document, you agree to all provisions of the copyright laws protecting it.The Cell-Averaging Constant False-Alarm Rate (CA-CFAR) algorithm was implemented and optimized in software on the NVIDIA Tesla C1060 GPU architecture for application in pulsed-Doppler radar signal processors. A systematic approach was followed to gradually explore opportunities for parallel execution and optimization by implementing the algorithm first in MATLAB (CPU), followed by native C (CPU) and finally NVIDIA CUDA (GPU) environments. Three techniques for implementing the CA-CFAR in software were identified and implemented, namely a naive technique, sliding window technique and a new variant which employs the Summed-Area Table (SAT) algorithm. The naive technique performed best on the GPU architecture. The SAT technique shows potential, especially for cases where very large CFAR windows are required. However, the results do not justify using the GPU architecture instead of the CPU architecture for this application when data transfer to and from the GPU is taken into consideration.enCell-Averaging Constant False-Alarm Rate (CFAR)DopplerRadarSignal processingConference PresentationVenter, C., Grobler, H., & AlMalki, K. (2011). Implementation of the CA-CFAR algorithm for pulsed-doppler radar on a GPU architecture. IEEE. http://hdl.handle.net/10204/5609Venter, CJ, H Grobler, and KA AlMalki. "Implementation of the CA-CFAR algorithm for pulsed-doppler radar on a GPU architecture." (2011): http://hdl.handle.net/10204/5609Venter C, Grobler H, AlMalki K, Implementation of the CA-CFAR algorithm for pulsed-doppler radar on a GPU architecture; IEEE; 2011. http://hdl.handle.net/10204/5609 .TY - Conference Presentation AU - Venter, CJ AU - Grobler, H AU - AlMalki, KA AB - The Cell-Averaging Constant False-Alarm Rate (CA-CFAR) algorithm was implemented and optimized in software on the NVIDIA Tesla C1060 GPU architecture for application in pulsed-Doppler radar signal processors. A systematic approach was followed to gradually explore opportunities for parallel execution and optimization by implementing the algorithm first in MATLAB (CPU), followed by native C (CPU) and finally NVIDIA CUDA (GPU) environments. Three techniques for implementing the CA-CFAR in software were identified and implemented, namely a naive technique, sliding window technique and a new variant which employs the Summed-Area Table (SAT) algorithm. The naive technique performed best on the GPU architecture. The SAT technique shows potential, especially for cases where very large CFAR windows are required. However, the results do not justify using the GPU architecture instead of the CPU architecture for this application when data transfer to and from the GPU is taken into consideration. DA - 2011-12 DB - ResearchSpace DP - CSIR KW - Cell-Averaging Constant False-Alarm Rate (CFAR) KW - Doppler KW - Radar KW - Signal processing LK - https://researchspace.csir.co.za PY - 2011 SM - 978-1-4577-1083-4 T1 - Implementation of the CA-CFAR algorithm for pulsed-doppler radar on a GPU architecture TI - Implementation of the CA-CFAR algorithm for pulsed-doppler radar on a GPU architecture UR - http://hdl.handle.net/10204/5609 ER -