Principal component analysis-based blind wideband spectrum sensing for cognitive radio

A principal component analysis-based blind wideband spectrum sensing (WSS) algorithm is presented, in which the WSS issue is transformed into a sequential binary hypothesis test under the framework of the general likelihood ratio test. The proposedmethod operates simultaneously over all the subbands rather than one single subband each time. Furthermore, the new method overcomes the noise uncertainty problem, and can also perform well without information about the channel, the primary signal, and the noise power. Most importantly, unlike the existing classical blind wideband detectors based on the information theoretic criterion, the decision threshold for the proposed detector can be flexibly determined according to the target false-alarm probability. Simulation results verify its effectiveness and superiority to the existing sensing algorithms.