The Research Of Cognitive Radio Cooperative Spectrum Sensing Technology

As we all know, the radio frequency spectrum resources, managed and allocatedby the state, are non-renewable. In recent years, the information industries of modernsociety are booming in an explosive speed, but the shortage of radio spectrumresources has gradually become the bottleneck in the development of the newcommunications technologies. Currently the widely used principle of fixed banddivision and distribution doesn’t make full use of the spectrum resources, thecognitive radio is proposed not only to solve the problem, but also improve theutilization of the spectrum. The most critical technology in cognitive radio isspectrum sensing, which requires the cognitive users to detect the idle band quicklyand access to communications accurately, and when the primary user signal apprears,it has to vacate the occupied band at same time. Cooperative spectrum sensing isdevolved on the basis of the single-node detection, it is greater used for its superiordetection performance and anti-fading. So deeply researchs on the cooperativespectrum sensing technology are done in this paper. The main contributions of thispaper can be outlined as follows：1. For minimizing the probability of false alarm subject to the constraint ofmissed detection probability, the paper proposes a new cooperative sensing modelwith an individual energy detection threshold for each SU on the basis of existingcooperative sensing technology. When the number of SUs increases, it may be hard tonumerically solve the optimization problem, then we develop a heuristic algorithm toachieve a suboptimal solution for the case of multiple SUs. Finally, the experimentalsimulations show the detection performance of the new scheme is better than theexisting method.2. For the detection of wideband signal, the wideband spectrum sensing schemeis also investigated in the paper and we propose a new method of widebandcooperative spectrum sensing based on multi-rate sub-Nyquist sampling (Co-MRSS) in order to reduce the number of signal samples and computational complexity. Eachcooperative node sends its sampling information to the fusion center and uses thecompressed sensing CoSaMP reconstruction algorithm to reconstruct the spectrumamplitude of the signal, then compares it with a detection threshold to decid e whetherthe primary user (PU) is existed. Simulations demonstrate that the proposed systemcan effectively sense the spectrum.