SBIR-STTR Award

During the past two years, RTA scientists have developed a numerical scheme for the detection and classification (D/C) of signals in highly noisy environments (negative SNR). This method performs D/C using nonlinear signal information, based on techniques developed from the theory of Chaotic dynamics. Preliminary tests on real-world data sets, recently supplied by Navy personnel, indicate that this method will likely outperform most current linear classification methods, provided the signal of interest contains some nonlinear correlations. In addition, this method has been modified to perform D/C on transient signals (e.g. active pulses), with correspondingly good performance. These algorithms are quite computationally efficient, and can be implemented on any standard workstation environment. The current aim of the research program will be to further refine these methods, measure operating characteristics, characterize performance on a variety of real-world data sets, and especially to modify the basic scheme for a variety of other problems of interest to the Navy. Anticipated benefits to the Navy will be electronic signal recognition and classification technologies with significantly improved performance characteristics, especially suited to high noise environments.

Keywords:
Detection/Classification Signal Processing Chaos Nonlinear Transients Nonlinear Dynamics Asw

The program objective is to develop a dynamical-systems based detector for low-SNR SONAR signals, which can provide significant gain over current energy detectors, on signal classes of high interest to the Navy. The dynamical detector will be bsed on RTA algorithms used in the Phase I program. We will investigate a variety of different model forms, to determine if they improve general performance, or can be spcifically tailored to signal classes of interest. We will investigate the usefulness of discrete dynamical mappings, higher-order polynomial fits, rational polynomial fits, transcendental functions, and the extended exponential auto-regressive model forms. We will develop a variety of improvements in the dynmaical gram code, the primary analysis tool for hte dynamical detection method, and contain its algorithmic embodiment to be developed into future onboard detection systems. The most important aspect of this program will be to demonstrate the improved dynamical detecitonmethods on real data sets of interest to the Navy. These data sets will be used to define ojective comparisons of the dynamical detector performance to an energy detector, as well as determine the signal classes on which the dynamical detector may prove most usefull.

Keywords:
Signal Processing Nonlinear Nonlinear Dynamics Acoustics Detection Sonar Roc Curves Asw