Fast and Accurate Signal Processing In Non-Gaussian Stable Environments

Robust Analysis has pioneered the development of extremely efficient algorithms for sonar/radar signal processing in heavy-tailed clutter, based on cutting-edge numerical methods derived from the theory of stable distributions. Robust Analysis is the developer of STABLE, the first commercially available software library for stable data processing. STABLE incorporates a Signal Processing Toolbox with comprehensive filtering solutions for impulsive signals. The goal of this SBIR is to extend the reach of the STABLE software to provide a comprehensive industrial-quality solution for sonar and radar engineering in heavy-tailed environments. The work during Phase I was focused on the design and implementation of speed optimizations able to accelerate the algorithms and enable real-time performance. Based on the significant performance improvements in stable methods for the processing of sonar and radar signals during Phase I, Phase II targets the demonstration of fast stable methods for the solution of estimation and image creation problems in sonar and radar applications. This demonstration will be done with a PC-based system that emulates the generation of data as done by sensors in their natural environments, and that further process the data with accelerated routines tuned for sonar/radar applications.