A Minimal Fuzzy Radial Basis Function Neural Network For Recognition Of Digitally Modulated Signals Of Multilevel Magnitude

Abstract

This work attempts to develop an intelligent decision support system for identification of digital signal type of multilevel magnitude based on fuzzy radial basis function (FRBF) neural network. This approach may solve the following problems: (1) time-consuming of learning in back-propagation neural network, (2) fluctuation of the values of parameters due to noise and fading (3) fuzzy linguistic-term judgment for signal classification. An FRBF network with fewer rules than classes to be discriminated is unable to recognize some classes, while, when the number of rules is increased up to the number of classes to be discriminated, a sharp increase in the performance is observed. Experimental results point out that the behavior of the FRBF network is closer to that of a competitive model showing a strong specialization of the fuzzy rules. Also the raw extracted features which form the input to the FRBF are weighted which enhances the effect of these features. By translating the importance of each feature using weights, the classifier leads to better results.