Automatic Classification of Radar Hybrid IntraPulse Modulation

In this paper, the automatic classification of radar hybrid intra-pulse modulations has been performed using spectrum-based characteristics in signal-to-noise ratio and using different methods in the field of modulation classification. First, the importance of combining radar modulations is discussed. Then the characteristics based on the signal spectrum and the behavior of each of these features for radar modulations are introduced. In the next step, the modulation classification methods for radar hybrid signals are examined and compared. At the end, the most optimal automatic classification method for combined modulations is selected and used. After 1000 trainings of each automatic classification method, the performance of the system improves to a great extent. This design also performs very well for very low SNRs and also offers much higher accuracy in that, unlike the previous methods, it shows effective efficiency in very low SNRs. The simulation result also confirms this method's very good performance in different conditions.