A robust method for automatic modulation recognition of linear frequency-modulated signals

Due to the differences in functions and applications, the bandwidth of the linear frequency modulation (LFM) signals adopted by radar radiation sources varies greatly. A single classification scheme cannot cover the identification of all bandwidth signals. This paper proposes a method for instantaneous frequency estimation based on short-time Fourier transform and phase difference combined with a robust classification decision tree, which realizes a robust automatic modulation recognition method for LFM signals. This scheme has been tested through simulation. Under the sampling rate of 2.4 GHz, this technical scheme can adapt to the automatic recognition of LFM signals with bandwidth ranging from 1 MHz to 1 GHz. This technical scheme significantly improves the robustness of automatic modulation recognition of LFM signals and can be used in engineering practice.