Spectrum Sensing for Cognitive Radio Based on Multiple Antennas

Spectrum sensing is a key component for enabling the cognitive radio paradigm. In this paper, we propose a novel totally-blind spectrum sensing technique for cognitive radio device equipped with multiple antennas, namely the Space Frequency Cross Product Sensing (SFCPS) algorithm. Existing correlation-based spectrum sensing techniques rely on the assumption that the received signals are correlated and their performance becomes poor when the signal correlation is low. By appropriately combining the received signals from multiple antennas, the proposed method creates new signals that are fully correlated and on which a sensing method is developed. SFCPS performs better than existing correlation-based techniques and with a lower computational complexity for small number of observed samples.