16 / 2019-12-02 02:10:00

A Frequency-domain Entropy-Based Compressive Sensing Algorithm

spectrum sensing; compressive sensing; modulated wideband converter; frequency-domain entropy

IEEE SAM 2020: Full Regular Papers > Compressed sensing and sparsity-based signal processing

Rooted in the compressed sensing theory, subNyquist spectrum sensing (SNSS) has been considered as an attractive approach to applying in wideband spectrum sensing in cognitive radio networks. Most existing SNSS algorithms require
the prior knowledge of number of PUs or the power of the noise, to determine the termination condition of the iterative recovery process. However, the number of PUs and the power of the noise are unknown and even changing in actual CR applications. With the help of entropy concept in information theory, a frequency-domain entropy-based sub-Nyquist spectrum sensing algorithm (FDEA) was proposed in this paper. The FDEA calculates the frequency-domain entropy of the recovery residual signal, which is compared with a pre-set gate γ. The merit of the proposed algorithm is that the pre-set gate γ is un-correlated with the power of the noise, so we do not have to estimate the SNR exactly. The simulation results illustrate that the FDEA if robust to different SNR.