Research On Cooperative Spectrum Sensing Algorithm Under Sub-Nyquist Sampling

With the rapid development of wireless communication technology,the increasing number of mobile terminal devices,and the increasing transmission rate of wireless communications,the demand for wireless spectrum resources has been increasing.However,the existing spectrum resources are scarce,and the reason for this problem is that the spectrum resources cannot be fully utilized.Therefore,how to make full use of the spectrum resources is one of the most critical problems to be solved in future communication network research.Cognitive Radio(CR)technology can effectively improve the upper limit of spectrum utilization because it can autonomously regulate the spectrum management mechanism.As a critical technology in CR,spectrum sensing includes Nyquist sampling-based and compressed sampling-based broadband spectrum sensing.However,while the former imposes a hardware burden and high communication overhead,the latter can recover signals from fewer measurements,thus reducing sampling pressure and enabling efficient data storage and transmission.Therefore,this thesis focuses on the improved matching tracking algorithm based on compressed sensing and the distributed collaborative compressed spectrum sensing algorithm based on the support set information,and the main contributions are as follows:(1)Since the inner product matching criterion in the reconstruction algorithm cannot accurately measure similar vectors,it will lead to a high misclassification rate of the best matching atoms,and the accumulation of interference terms in the iterative process will also affect the reconstruction accuracy of the algorithm,which cannot meet the demand of higher precision data reconstruction.To address this problem,a secondary screening-selective backtracking matching pursuit based on Dice matching(DSS-SBMP)algorithm is proposed in this thesis,which introduces the matching criterion of Dice coefficients to solve the problem that the matching criterion of inner product is inaccurate in measuring the similarity between two vectors,reduces the number of incorrect indices in the support set corresponding to the atoms by the secondary screening of the atoms,and introduces selective backtracking to overcome the phenomenon of excessive backtracking in the iterative process.Simulation results show that the algorithm can retain more correct atoms in the iterative process,and it has better reconstruction performance and good application value compared with similar greedy algorithms.(2)Since single-user spectrum sensing can no longer meet the access requirements of large-scale user device deployment,cooperative spectrum sensing can improve spectrum sensing performance by utilizing the spatial diversity gain of multi-user collaboration.In view of this,this thesis proposes a distributed collaborative spectrum sensing algorithm based on the support set fusion by applying the DSS-SBMP algorithm to the local spectrum sensing of cooperative cognitive users as a way to obtain support set estimation information.Using the joint sparsity between signals,the support set estimation information is fused using the average consensus technique,and the weight vector is introduced as an a priori condition to make the cognitive users reconfigure locally again until the algorithm converges.The effectiveness and reliability of the proposed algorithm are verified by simulation experiments,and the simulation results show that the proposed algorithm has better detection performance compared with the traditional collaborative spectrum algorithm.