Cyclostationary-based cooperative compressed wideband spectrum sensing in cognitive radio networks

In this paper, a cooperative cyclostationary compressed spectrum sensing algorithm is proposed to enable accurate, reliable and fast sensing of wideband spectrum. In the proposed algorithm each secondary-user (SU) sends the compressed data vector to the fusion center (FC) which has a copy of the sensing matrices for all cooperated SUs. Then, at the FC, the fast fourier transform accumulation method (FAM) based on cooperative multitask compressive sensing (MCS) algorithm is employed to recover the spectral correlation function (SCF) from the compressed measurements. The proposed algorithm has two main components. The first component exploits the cooperation between SUs to produce an estimate of the investigated signal spectrum using multi-task compressive sensing. In the second component, the cyclic feature detection is performed based on the recovered SCF function. Simulation results demonstrate the robustness and the effectiveness of the proposed framework against both sampling rate reduction and noise uncertainty.