Classification of radio and telecommunications terminal equipments with deriving radiated emission wave patterns

At the present day many engineering applications are tending to pattern recognition systems. Pattern recognition is definition and classification of complex symbol samples and objects which have similar attributes points and be associated through specified attributes and characters. Pattern recognition is performed in three steps as preprocess, attribute derivation and classification. The most important step of pattern recognition is attribute derivation. Because, the most important factor in determining classifier is determination of the right attributes.For this study, radiated emission wave symbols of radio and telecommunications terminal equipments are derived from Market Surveillance Laboratory which is located within Information and Communication Technologies Authority. Attribute derivation technique which is based on ripple packet transformation time frequency representation entropies was applied to these symbols. Classification is made by giving the obtained attributes to the back propagation artificial neural networks. Accurate recognition performance of the system which is aimed to determine the equipment type was % 92. It is considered that using this method as decision maker and assistant especially in telecommunication area will improve the reliability.