An Algorithm for Analog Modulation Classification
Research in automatic modulation classification systems has been carried out for a long period of time and many methods and classifiers are presented for both analog and digital modulations. The purpose of such systems is to properly recognize the unknown signal’s mod-ulation type and transfer it to the demodulator in order to fully recover the modulating signal. The program code developed in LabVIEW graphical programming environment represents a new technical solution of the analog modulation classification problem. It enables the usage of fast-operating hardware for real-time testing. The description of automatic modulation recogni-tion algorithm for analog modulations programmed in LabVIEW is presented in this article. Among analog modulations especially double sideband (DSB), single sideband (SSB) and fre-quency modulated (FM) signals have been considered.
The LabVIEW code consists of three parts, each maintaining specific tasks. The first part has to implement the task of a receiver and process, the received signal and extract key fea-tures, but here only simulated signals have been considered. The second part represents the algorithm of the classifier and the third part is the demodulator block. After simulating a signal and extracting the key features, their values are passed to the classifier’s inputs. An artificial neural network was chosen as a classifier for this task. The network implements self-learning before testing and based on the learning results, classifies the current signal modulation type.
Afterwards, testing was carried out, and the results are shown in this paper.