Communication signal modulation identification method based on BP neural network

The invention provides a communication signal modulation identification method based on a BP neural network, and belongs to the field of communication signal processing, and the method comprises the steps: obtaining original modulation signals which comprise six kinds of signals: 2ASK, 4ASK, 2PSK, 4PSK, 2FSK, and 4FSK; performing segmentation and noise reduction on the original modulation signal by using a signal preprocessing module; extracting feature parameters according to the processing result of the last step, constructing a multi-dimensional vector to serve as a training set, wherein the feature parameters comprise five feature parameters based on instantaneous statistics, namely gamma max, sigma aa, sigma af, sigma ap and sigma dp; building a neural network model; reconstructing a signal and a characteristic parameter vector as a test set, and performing signal identification by using the test set; simulating the recognition effect of the test set; classifying and identifying each signal; and calculat