A cooperative spectrum sensing algorithm based on unsupervised learning

Spectrum sensing is an essential problem in cognitive radio and has been discussed a lot in recent years. In this paper, a cooperative sensing algorithm based on unsupervised learning is proposed. The unsupervised learning framework that does not require the training data to be labeled, including K-means clustering and Gaussian mixture model, is introduced into the scheme. The more robust features including eigenvector and eigenvalue are fed into the classifier. Simulation results are shown to illustrate the performance of the proposed algorithm.