Spectrum Sensing for Cognitive OFDM Over Line-of-Sight Doubly Selective Fading Channels

Spectrum sensing for fast-moving vehicles in cognitive radios over a general line-of-sight (LOS) doubly selective fading channel, which is a practical environment for modern mobile orthogonal frequency-division multiplexing communications, should be addressed. However, time- and frequency-selectivity inherent in radio channels is a limiting factor for the detection performance. By discovering the complement property of correlation profile (introduced by the multipath delay spread) and adopting the proposed estimation for the correlation function of LOS channels (introduced by the mobility), the decision metric can be transformed to the ideal one without time- and frequency-selectivity. As such, the detection performance can be greatly enhanced by the proposed technique without needing any channel equalization techniques on received signals.