Quickest Spectrum Sensing With Multiple Antennas in Rician and Correlated Rayleigh Channels

摘要

In this letter, we study the quickest spectrum sensing employing multiple receive antennas at the cognitive user when the received signal experiences Rician and correlated Rayleigh channels. For both channels, we derive the joint probability density functions required to form the optimal cumulative sum (CUSUM) detector. We explore the effects of the Rician $K$-factor or correlation on the performance of the quickest spectrum sensing. Mismatched channel conditions are also considered, and we show that the sensing performance at a particular correlation coefficient or the Rician $K$-factor heavily depends on the true channel and is relatively insensitive to the detector.