Self-correlation and deep learning combined communication signal rapid detection method and device

The invention relates to an autocorrelation and deep learning combined communication signal rapid detection method and device. The method includes generating communication signals having different modulation patterns using software radio, and receiving the communication signals by a receiver; performing segmentation processing and labeling processing on an observation signal of the communication signal received by a receiver to obtain a segmented signal sample with a label; calculating a self-correlation result of the segmented signal sample according to a self-correlation function, constructing a training set according to the self-correlation result and the label, and training a preset deep learning model according to the training set, so that the deep learning model can judge whether a signal exists or not; and carrying out segmentation processing on the observation signal to be detected, and inputting the observation signal subjected to segmentation processing into the trained deep learning model to obtain