Detection of primary user signal in wideband cognitive radio networks exploiting DCT as sensing matrix

Wideband spectrum sensing is a challenging task in wideband cognitive radio networks. It needs to be efficient, robust and fast. However, there are many challenges facing sensing in wideband spectrum. One of these challenges is Nyquist sampling rate bottleneck which required high speed DSP and large storage spaces. Compressive sensing is a pioneer solution for wideband spectrum sensing which have proved sampling below Nyquist criterion. In this paper, we provide mathematical expression that computes the number of measurements required for compressed detection exploiting DCT as sensing matrix. The simulation results show that the accuracy of detection for certain number measurements is similar to the results obtained by compressed measurements based spectrum sensing for wideband cognitive radio systems with small relative error. The comparison is based on term of probability of detection versus compression ratios.