Research On Wideband Spectrum Sensing Technology In Cognitive Radio Networks

With the development of wireless communication services,the demand for wireless spectrum resources is increasing.Therefore,the requirements of spectrum sensing in cognitive radio networks also extend to the detection of spectrum access opportunities over a wider frequency band.According to the sampling rate,wideband spectrum sensing can be divided into two categories: wideband sensing based on Nyquist sampling and wideband sensing based on compressed sampling.The former refers to the signal sampling rate not less than the Nyquist sampling rate;the latter refers to the signal sampling rate much lower than the Nyquist sampling rate.This thesis studies the wideband spectrum sensing methods based on compressed sampling.In view of the high complexity of the reconstructed compressed sensing algorithms,the influence of the measurement matrix selection on the performance of the compression sensing,and the influence of the cognitive users' positions on the spectrum sensing decision-making results,efficient and reasonable algorithms are designed according to the characteristics of broadband spectrum sensing signal to reduce the computational complexity and improve the robustness to channel noise.The main works and contributions are as follows:1.Aiming at the high complexity of the reconstructed compressed sensing algorithm and the drawback that the existing compression detectors are only applicable to the known signals with known parameters,a non-reconstructed compressed wideband spectrum sensing algorithm is proposed for the known signals with unknown parameters.Based on the generalized likelihood ratio criterion,the theoretical detection performance bounds of time-varying amplitude signals are derived and the computational complexity is analyzed.The simulations are carried out under Rayleigh and Rician wireless fading channel conditions respectively,and the effects of different parameters on the performance of the detector are studied.The simulation results show that the proposed algorithm not only avoids the high computational complexity of the reconstructed compressed sensing algorithms,but also has a higher detection probability for the wideband spectrum sensing signal.2.Aiming at the problem that the rationality of measurement matrix design directly affects the signal sampling results,and then affects the detection performance of non-reconstructed compressed spectrum detection,a wideband spectrum sensing algorithm using optimized measurement matrix is proposed.The proposed algorithm takes the F norm of the difference between the Gram matrix and the approximating equiangular tight frame matrix as the objective function,and optimizes the measurement matrix by iterative approach.The wideband spectral sensing algorithm with Gaussian random matrix and the proposed one with an optimized measurement matrix are compared.The simulation results show that compared with the wideband spectrum sensing algorithm with Gaussian random matrices,the proposed algorithm with the optimized measurement matrix has the less correlation between sparse basis and measurement matrix,so that it can obtain a better detection performance.3.Aiming at the fact that the received signals of the broadband cooperative spectrum sensing decisions may be seriously interfered or faded due to various factors,such as,topography,landform,building structure or dense distribution,which results in the unreliable of the received broadband sensing results,an optimal fusion rule considering the reliability of cognitive users' location is proposed.According to the space-time state of a certain cognitive user's environment,a low or high trust value is assigned.When a cognitive user moves to a new location,a new trust value that reflects the current location is assigned again and is used as a basis for decision fusion of the received signal.The proposed compressed sensing scheme based on location reliability and the traditional hard combination decision one are compared.The simulation results show that the proposed scheme has better detection performance than the traditional cooperative spectrum sensing algorithm in high-shadow areas,and it also meets the requirements of cognitive radio network for the reliability of sensing results in urban environments.