SBIR-STTR Award

Neural networks as well as classical transformation techniques have been successful in the field of static pattern recognition. However, the problem of reliably detecting and recognizing weak, temporally varying signals remains a challenge. New techniques are needed to detect and recognize conventional signal types within a dense, dynamically, changing, interference environment. Given their successful application to the field of static pattern recognition, neural networks offer a promising approach to solving this difficult problem. The approach for detecting and recognizing weak, temporal Signals of Interest (SOIs) and Signals Not of Interest (NSOIs) consists of three distinct parts: algorithm development, simulation environment development, and algorithm validation. These three parts are designed to achieve the stated objectives by providing careful analyses of existing neural network architectures as well as the development of hybrid architectures and by thoroughly validating these algorithms in a controlled, reproducible, simulated environment. This research is expected to produce a set of algorithms which can be applied to separating SOIs and SNOIs, as well as a set of heuristics for selecting which algorithms to apply given characteristics of the type of signal being analyzed. This project deals with the three major parts of research which address algorithm development, signal simulation environment, and algorithm validation.

Benefit:
The proposed project has tremendous potential use by the Federal Government for such signal recognition and classification tasks as aircraft recognition, voice recognition, evaluation of radar returns, and various electronic intelligence applications. The modular architectures support implementation in a compact workstation environment which will provide easy portability of the entire system.

Keywords:
temporal, temporal, transforms, Wavelets, Neural networks, Self-organizing Networks, weak, Temporal Pattern Signal Recognition, Classification, recognition/detection, Gabor, signal, Signals

The Phase I research successfully developed and validated an integrated methodology of neural network (NN) algorithms combined with feature vector generators for the automated detection and recognition of weak, temporal signals or interest (SOIs) and signals not of interest (SNOIs) in the presence of non-Gausian noise. This methodology successfully identified all weak SOIs without any prior knowledge of the SOI's presence in the signal for all cases tested. The Phase II effort is expected to better identify weak SOIs with one or more SNOIs by combining this research team's single network/feature vector combinations with multiple features and hybrid network approaches. Phase II will focus on developing: a) nerual networks with multiple input features, b) hybrid neural netowrks, and c) extension of feature extraction work including the preliminary wavelet work and other feature extractors.

Benefit:
The results of this research will form the basis for developing a commercial product which can be used by the communications industry for recognizing and detecting weak temporal signals of interest.

Keywords:
Self-organizing Networks, temporal, signal, Wavelets, recognition/detection, Classification, Temporal Pattern Signal Recognition, weak, Gabor, Neural networks, Signals, transforms