Harnessing the Power Of GPUS for Nisar Data Simulation & Processing

The NASA-ISRO Synthetic Aperture Radar (NISAR) mission will be instrumental in studying the entire global landmass for various themes like ecosystem, cryosphere, solid Earth, ocean applications, etc. Due to the high-resolution wide-swath leading to the large volume of NISAR data, the traditional CPU computing was found insufficient for meeting the turnaround time for generating various levels of data products. In this paper, we explore the use of Graphics Processing Units (GPUs) to accelerate data processing. Primarily, we focus on four modules: echo data simulation, geolocation grid generation, amplitude correlation, and resampling. The GPU implementation of these modules outperforms the CPU processing and provides a significant speed-up.