Wideband Spectrum Sensing Using Compressive Sampling Based Energy Reconstruction

In this paper we perform wideband spectrum sensing employing a Generalized Likelihood Ratio (GLR) detector which uses the sub-channel energies obtained by a Compressive Sensing (CS) based method. The wideband spectrum sensing, in high utilization regimes, improves access to the vacant sub-channels in a cognitive radio network. The existing GLR detector requires the energy in each sub-channel acquired by a filter bank which imposes high Computational Complexity (CC) to satisfy high resolution. Here, we apply the received wideband signal to several appropriate wideband filters and employ CS algorithms to obtain the sub-channel energies. This technique was used to reconstruct energy in each sub-channel, earlier; however, in this paper, we employ it for wideband spectrum sensing. We also decrease the CC by extending the CS based method to the case that one primary user occupies several sub-channels. Simulations compare the performance and the CC of our proposed algorithm against existing methods.