Research On Spectrum Sensing In Cognitive Radio Networks

Along with the scarcity of usable spectrum, cognitive radio, which is intelligence software defined radio,is regarded as the most promising technology to improve the spectrum efficiency. In cognitive radio networks, secondary users(SUs) can access the channels that belong to primary users(PUs) when PUs is not using the channels by spectrum sensing technology.This article is main concerned on spectrum sensing. The full duplex is implemented in recent years, which brings new opportunities for wireless networks. For example, full duplex can be used in cognitive radio networks, in which SUs can transmit the data and sense the channel at the same time. Using previous full duplex spectrum sensing SUs have to stop transmitting data when the PUs access the channel, SUs can’t transmit data until find another idle channel, which will decrease the spectrum usage efficiency. Our proposed method, called asymmetric full duplex spectrum sensing, can monitor the currently-used channel and its adjacent channels at the same time, so SUs can hop to adjacent idle channel when the PUs access the currently-used channel immediately. Result from the high sample rate of AFDSS, we use compressive sampling to instead of the traditional Nyquist sample. Previous works on compressive sensing mostly focus on sensing accuracy but fail to take time-efficiency into consideration. We proposed to use Structurally Random Matrix(SRM) instead of Gaussian Random Matrix(GRM) as compressive matrix to improve the speed of compressive sensing.In section five, we conduct simulation to verify the availability and efficiency of our proposed AFDSS and TECSS. The result show that networks throughput improved mostly by 200% when SUs use AFDSS compared with using traditional full duplex sensing. And the accuracy of SRM is not decreased or slightly decreased, but the sensing speed is improved by 100%.