Parallel Implementation of the Range-Doppler Radar Processing on a GPU Architecture

Graphic processing units (GPUs) is widely used to accelerate the processing speed of the radar detection procedure, including the range compression, coherent integration and constant false alarm rate. Specifically, detailed parallel design of the radar algorithm and the thread programming are shown. The experimental results show that, by engaging the parallel technology into the radar processing procedure, much high speedup ratio can be obtained. Furthermore, precise target detection can be guaranteed.