Research On Cooperative Spectrum Sensing Algorithm In Cognitive Radio System

With the development of wireless communication,radio frequency spectrum resources are increasingly tense,and the fixed spectrum allocation policy can not meet the current demand for spectrum resources.Therefore,cognitive radio technology came into being,which enables effective reuse of licensed spectrum by changing operating parameters dynamically.Spectrum sensing technology is the foundation and premise of cognitive radio networks.It effectively allocates spectrum to unlicensed users by detecting the spectral holes that are not occupied by the primary user,so as to alleviate the problem of spectrum resource shortage effectively.Because of the simplicity and its lack of prior knowledge of primary users,energy detection technology has become the preferred method for cognitive radio spectrum sensing.In practical communication networks,such issues as multipath fading,shadowing and receiver uncertainty can affect the performance of spectrum sensing seriously.However,multi-user cooperative spectrum sensing techniques can reduce the interference to cognitive radio systems effectively caused by the above factors.The traditional energy detection technology is based on the judgment of the fixed threshold,which is very sensitive to the noise uncertainty under the actual situation.Therefore,to solve this problem,this thesis presents an improved energy detection algorithm.Predict the current activity of the primary user by storing the state information of the primary user in the continuous period.Based on the prediction results,by introducing the noise uncertainty coefficient,the threshold is switched dynamically to improve the detection probability,reduce the false alarm probability and enhance the robustness of the system to the SNR wall.To improve the fusion technology of the cooperative spectrum sensing,this thesis combines the hard and soft judgments program.Majority rules in hard-decision techniques are combined with maximum-ratio-merging techniques in soft-decision techniques.After the secondary user completes the local sensing by using the improved energy detection technology,the decision result and the perceived signal energy are stored as a binary array and sent to the fusion center.The fusion center will carry out joint arbitration according to the soft and hard judgment plan,and will finish the final judgment when the two judgments are the same.The proposed algorithm effectively makes up for the systematic error caused by single decision,reduces the impact of noise uncertainty on the system and improves the effectiveness of the detection system.The Monte Carlo simulation results show that the proposed schemes can effectively reduce the influence of SNR Wall and improve the spectrum utilization rate,so as to improve the performance of spectrum detection.