Research On Wideband Spectrum Sensing Methods Based On Sub-Nyquist Sampling

The rapid development of wireless communication technology and the increasing demand of wireless communication services have resulted in the scarcity of spectrum resources.Traditional wireless spectrum resource adopts static spectrum allocation strategy,which cannot make full use of limited spectrum resources.In order to overcome this drawback,based on the concept of dynamic spectrum sharing,researchers have proposed the cognitive radio system,which can greatly improve the utilization of spectrum resources.Spectrum sensing technology is one of the most important basic technologies in cognitive radio system.There are many narrowband spectrum sensing methods which own mature technologies.The difference between narrowband and wideband signals however makes it difficult to apply narrowband methods to wideband spectrum sensing.The innovative wireless communication technology extends the detection range of cognitive radio from the original hundreds of MHz to up to thousands of MHz in the combined frequency of multiple authorized communication systems.In this situation,investigating wideband spectrum sensing has become an important way to alleviate the shortage of spectrum resources.This thesis has studied the traditional wideband spectrum sensing methods based on Nyquist sampling.This type of methods divide the wideband into a series of continuous sub-bands which are not overlapped,and each sub-band is detected by narrowband sensing technology,so as to realize the wideband spectrum sensing.Through the research and analysis of multiband joint detection approach spectrum sensing and filter bank spectrum sensing approach for cognitive radios,the advantages and disadvantages of these methods are compared,and the sampling bottleneck problem faced by these methods in large bandwidth signal sensing is analyzed,which highlights the significance of research on wideband spectrum sensing method based on sub-Nyquist sampling.This thesis has also studied the compressed sensing theory and its application in the field of wideband spectrum sensing based on sub-Nyquist sampling.This class of algorithms mainly use the inherent sparsity of wideband signal to realize wideband spectrum sensing at a sampling rate much lower than Nyquist rate.In this thesis,three typical sparse signal reconstruction algorithms are used to reconstruct the spectrum of wideband signals,and the performance of the algorithms is compared and analyzed according to the simulation experiments.In this thesis,the fast compressed power spectrum estimation algorithm has been studied,simulated,and compared with other algorithms.The wideband spectrum sensing methods based on compressed sensing theory often require the wideband signal to be sparse,and are not robust to noise.In order to solve these problems,a new sampling structure and a multi-antenna assisted wideband spectrum sensing algorithm are proposed in this thesis.Instead of directly reconstructing the wideband signal,the new algorithm chooses to reconstruct the power spectrum of the wideband signal.By analyzing and utilizing the relationship between the compressed samples and the statistics of the wideband signal,the new algorithm reconstructs the autocorrelation matrix of the wideband signal and then realizes wideband spectrum sensing.Simulation results show that the proposed algorithm can achieve good sensing performance in low SNR environment,and is suitable for non-sparse wideband signals.