Comparative Studies fo SAR and SAL ATR Performance on the DoD HPCs
Ross, Jacob (Etegent Technologies LTD)
Intersection of Digital Engineering and High Performance Computing/High End Computing
This study presents selected results of a comparative analysis of ATR performance on SAR and SAL imagery over a wide swath operating conditions (OCs). The goal of the study is to identify how SAL performance will deviate from the performance trends in the SAR space. While SAR has a rich history in the ATR domain; SAL is relatively unexplored. This study will aid in sensor and ATR design for the SAL and ATR practitioner.
Due to the lack of availability of measured SAL data, synthetic data plays a crucial role in this study. High performance computers (HPCs) enable a deep dive on performance prediction with synthetic data. Leveraging HPC resources enabled the analysis of all possible combinations of pertinent OCs as identified in the study. Additionally, HPCs allow the rapid generation of massive amount of imagery to encompass the explored OC space.
Techniques for SAL and SAR data generation, algorithm development, and ATR evaluation on the DoD HPC systems are presented in addition to key findings from the ATR study.