Research On Wideband Spectrum Sensing For Cognitive Radio

With the emergence of new technologies such as the Internet of Things,cloud computing,and artificial intelligence,the demand for wireless access is growing rapidly.The static spectrum management method is no longer effective enough,leading to an increasingly obvious contradiction,that is,a large number of newly emerging nodes cannot obtain suitable wireless access frequency bands and most of the allocated spectrum utilization rate is low.Cognitive radio(CR)utilizes dynamic spectrum access strategies to provide a large number of access opportunities for unlicensed secondary users.Spectrum sensing is the premise of cognitive radio and is responsible for detecting the activity status of the primary user(authorized user).Furthermore,wideband spectrum sensing technology can detect a wider spectrum space at one time and has been widely studied.A wider detection interval means a higher Nyquist sampling rate,which brings a huge computational and storage burden to the sensing system.Based on the sparse nature of wideband signals,compressed sensing technology can accurately reconstruct the original signal by compressed sample that can be obtained by sub-Nyquist sampling,which providing a new paradigm to acquire information.At present,the mainstream sub-Nyquist sampling structure in the field of wideband spectrum sensing includes analog information converter(AIC),multi-coset sampling(MCS),and modulated wideband converter(MWC).However,most reconstruction algorithms do not fully consider the characteristic changes brought about by these three sampling structures,resulting in high computational complexity and unsatisfactory perception accuracy.Based on the above problems,this article focuses on the research of reconstruction algorithms,and the main contributions are as follows:(1)In view of the high computational complexity of the reconstruction algorithm and low perception time efficiency,a new reconstruction algorithm,called nearest orthogonal matching pursuit(N-OMP),is proposed based on the modulated wideband converter.The algorithm utilizes the power spectrum slicing feature caused by pseudo-random sequence and low-pass filtering to simplify the reconstruction process.The specific method is to calculate the correlation coefficient between the residual matrix and the column vectors corresponding to the two adjacent subbands after detecting that a certain subband is occupied,and then compare the correlation coefficients to directly judge the occupation of the two adjacent subbands,for reducing the number of iterations of the reconstruction algorithm.Furthermore,this paper derives the spectrum occupancy discriminant based on the correlation coefficient and gives its proof process.Theoretical derivation and simulation experiments prove that,compared with the orthogonal matching pursuit algorithm,the proposed algorithm can reduce the computational complexity by up to 50% while showing better support reconstruction accuracy.(2)In view of the low performance of the traditional sampling scheme in the wideband domain and the unsatisfactory detection accuracy of compressed spectrum sensing using the sparse characteristics when the sparse level is relatively high,the compressed sampling reconstruction scheme based on the MWC is adopted.The reasons for partial missed detection in the restored signal support set were analyzed.Based on the support information recovered by a single node,a binary decision model suitable for wideband detection is established.Furthermore,utilizing the idea of space diversity,a centralized cooperative wideband spectrum sensing algorithm using hard fusion is proposed to improve the overall support set recovery accuracy.The simulation results confirmed that the algorithm has significantly improved the accuracy of the support set recovery,while showing strong robustness,which can provide reliable support for IoT network communication.