Research On Radio Frequency Spectrum Detection Technology Based On Multidimensional Spaces

With the application and development of wireless communication technologies, spectrum resource relative to the increasing communication demand becomes very limited. Spectrum detection technology is the key technology to find spectrum holes and enhance the utilization ratio of spectrum resource. The spectrum holes exist in multiple dimensions correspond to the multi-dimensional signal transmission. Frequency is the inherent dimension, and other two basic dimensions refer to time and space. This dissertation focuses on the spectrum detection of multiple dimensions in cognitive radio networks which includes the spectrum detection in time dimension, space dimension and joint time-space dimensions. First the optimal threshold of energy detection for spectrum sensing in cognitive radio (CR) networks, a cooperative spectrum sensing method based on dempster-shafer (D-S) theory of evidence in cognitive radio networks, and the double-threshold energy detection method of spectrum sensing for cognitive radio networks under noise uncertainty environment are studied in the time dimension. Then a spectrum holes detection method for cognitive radio networks applied in typical scenarios, and a space-time two-dimensional spectrum sensing method using mean detection probability constraint in cognitive radio networks are studied in the space-time dimensions.The main contributions of this dissertation are as follows:1. For the energy detection in the spectrum sensing, a local optimal detection threshold and the near optimal threshold of the half voting cooperative spectrum detection are derived. Analysis and numerical results show that the system error probability which refers to the sum of false alarm probability and missed detection probability is minimized for single user detection with the local optimal threshold, and the system error probability is close to the minimum value when the near optimal threshold is used for the multiuser cooperative detection with half voting rule under low signal noise ratio (SNR).2. For the cooperative spectrum sensing, a spectrum detection method is proposed based on D-S theory of evidence, which modifies the conflicting evidence source and combination equation. Analysis and simulation results show that the proposed method can further improve the detection performance in cognitive radio networks when compared with the "And", "Or" fusion rule and other D-S theory of evidence fusion methods and can be used as a general spectrum sensing method in additive white gauss noise (AWGN) and Rayleigh fading channel.3. For the energy detection in the spectrum sensing under noise uncertain environment, a double-threshold detection method is proposed. A high threshold and a low threshold are set according to noise uncertain range in the method and the test result is judged only when it falls outside the range of two thresholds. Analysis and simulation results show that the proposed method can overcome the influence of noise uncertainty and the detection performance under noise uncertain environment is even better than the one under noise certain environment with the accurate noise power. The detection performance in AWGN and Rayleigh channels are obviously improved with the proposed method.4. For the several typical scenarios in the cognitive radio networks, a spectrum holes detection method is proposed. By removing the influence of the space false alarm (SFA), the spectrum holes detection probability of single cognitive radio (secondary user) under path loss and Rayleigh fading is derived. Meanwhile the spectrum holes detection probability of multi-users cooperative detection and the one of single-user detection in multi-bands under path loss are derived. Analysis and simulation results show that the spectrum holes detection probability is related to the sampling times. The spectrum holes detection probability of the multi-users cooperative detection is not definitely better than that of single-user detection. The spectrum holes detection probability will be further decreased in Rayleigh fading but greatly increased in multi-bands.5. For the spectrum holes detection in cognitive radio networks, a space-time two-dimensional sensing method which uses mean detection probability constraint is proposed. The constraint for secondary user to access the spectrum of authorized user (primary user) is satisfied by deriving the mean detection probability within the whole detection range, based on which the detection threshold is set to derive the spectrum holes accessing probability in time, space and space-time two dimensions theoretically. Analysis and numerical results show that the false alarm probability in time and space dimensions are both decreased and the spectrum holes accessing probability in space-time two dimensions is greatly increased with the proposed method.