GMTI tracking improved by 18 dB using cognitive algorithm

Abstract

Existing tracking algorithms face combinatorial complexity in heavy clutter. Their performance is limited by the number of computer operations, they do not extract all the information available in radar signals, and do not reach Cramer-Rao performance bounds. A cognitively inspired algorithm was developed and applied for improved tracking. Models for GMTI tracks have been developed as well as cognitive architecture incorporating these models. The cognitive tracker overcomes combinatorial complexity of tracking in highly-cluttered scenarios; its performance achieves Cramer-Rao Bounds and results in about 20 dB (two orders of magnitude) improvement in signal-to-clutter ratio.