Enhancing electronic combat system digital signal processing using neural networks

The study focuses on the use of neural networks in three distinct areas of electronic combat (EC) signal processing: pulse sorting, pulse repetition interval (PRI) analysis, and emitter identification. Results of testing using neural networks show that neural networks have the potential to outperform conventional software in many cases. Pulse sorting is especially promising due to the way in which the Kohonen self-organizing map sorts pulses with ambiguous pulsewidth or frequency values. PRI analysis yields good results even when pulses have been dropped or in the presence of noise. It has been shown that the conventional software used for preprocessing PRI analysis input data is critical to the success of the network. While neural networks may never totally supplant conventional EC software, they can certainly be used effectively to enhance EC systems.<>