Usage of production-based expert system and neural network for signal recognition

Abstract

At the current paper the system of signal type automatic recognition for the problem of automated testing of radar stations based on collaborative usage of production system and neural network set is presented. The main approaches for solving the problem of automated recognition digital and analog modulation types are shortly outlined. Characteristics calculated based on time-domain and frequency-domain signal representation, typical for the radar impulses are considered. The structural schema of the software for automated classification of radar impulses is presented.