ADVANCED SIGNAL WAVEFORM CLASSIFIER

The rapidly changing military signal environment has reduced the probability of detecting and correctly identifying signal emitters with conventional signal analysis techniques. Increasingly complex waveforms covering large spectral band-widths are now being used. In addition, the number of active emitters has increased dramatically. Coupled with this rapidly rising processing load is the use of techniques designed to counter the signal database identification process; for example intentionally increasing the parameter limits of the emitter or changing one or more of the emitter operating parameters. Clearly new signal processing techniques are needed for both emitter detection and identification. Our approach addresses both of these problems with an optical signal processor/receiver for emitter detection and angle-of-arrival determination followed by a neural network uses a unique classifier scheme already proven to work well in EW pulse classifier applications, and optical emitter detection and identification applications.