Multiantenna detection of multicarrier primary signals exploiting spectral a priori information

Abstract

We consider the problem of detecting a primary signal over a wireless channel by a multiantenna cognitive spectral monitor with knowledge of the spectral shape of primary transmissions employing multicarrier modulation. As a starting point, a Locally Most Powerful (LMP) test is derived for the single-antenna case. The asymptotic performance improvement of the LMP detector over the standard Energy Detector (ED) is quantified in terms of the primary spectral mask. For the case of an ideal bandpass mask the LMP test is Uniformly Most Powerful. With multiple antennas, optimal detectors require channel knowledge and therefore are not well suited to practical implementation. We propose a realizable multiantenna Incoherent Detector, and compare its performance with that of the ED and LMP single-antenna tests under both deterministic and Rayleigh slow-fading channel scenarios. In the latter setting, multiple antennas introduce detection diversity that determines the slope of the probability of miss curve, and thus the overall detection performance.