Residual Correlation Matrix Detection Based Blind Sub-Nyquist Spectrum Sensing for Cognitive Radio Networks

Benefiting from compressed sensing theory, sub-Nyquist spectrum sensing (SNSS) has been considered as a promising way to deal with the implementation limitations of conventional wideband spectrum sensing in cognitive radio (CR) networks. However, in most existing SNSS methods, the prior knowledge of the monitored frequency bands is needed to determine the termination condition of the iteration process in implicit signal recovery stage, which may be difficult to acquire in practical CR scenarios. To address this dilemma, a blind SNSS algorithm for multicoset sub-Nyquist sampling framework, referred to as Residual corrElation mAtrix Detection (READ), is proposed to control the iteration process autonomously and appropriately. Simulation results show that, without any prior knowledge, the READ algorithm can precisely determine the support of a multiband signal contaminated by the background noise in a wide range of signal to noise ratio (SNR).