Spectrum sensing method based on unsupervised machine learning classification algorithm

The invention relates to a spectrum sensing method based on an unsupervised machine learning classification algorithm, and belongs to the technical field of wireless communication. According to the method, a geographic area in a target network is divided into Q grids in the same size, and the whole scheme implementation process is divided into four stages. In the first stage, a PUT transmission mode classifier TM1-Classifier is obtained; in the second stage, a grid label classifier TM2-Classifier is obtained; the third stage is to obtain a PUT transmission mode label at the current m moment; and the fourth stage is to obtain an LFB access indication label corresponding to the grid. The invention is based on a convolutional neural network in machine learning and a spectrum sensing scheme ofa threshold detection algorithm; under the condition that the PUT position is unknown, space-time idle spectrum resources are flexibly allocated, the opportunity of accessing the LFB is increased, and therefore the spectrum