Classification and model reconstruction method of non-cooperative multifunctional radar waveform unit

There are many variations in the form of multifunctional radar signals. At present, the problem associated with the highest concern in the field of electronic reconnaissance is the inversion of its working mode from the reconnaissance data. A method is proposed for waveform unit classification and model reconstruction using the non-cooperatively extracted waveform units as input data based on the hierarchical structure of multifunctional radars. The method performs multiparametric matching on waveform units subject to the constraint of the match level to obtain the similarity measure between waveform units; it then recursively expands the waveform unit pairing score matrix to complete the classification of the waveform units. Further, based on similar waveform units, the matching state matrix of the waveform unit model can be reconstructed. The feasibility and performance of the proposed method are verified by setting up typical multitype simulation experiments. The experimental results show that the proposed method can complete the classification of waveform units in the small-data content and extensive noise intensity cases and supports the reconstruction of waveform unit models with low errors.