Research On Multi-modal Fusion Spectrum Sensing Technology

With the rapid development of wireless communication services and the popularization of mobile devices,the available spectrum resources and the demand for communication services have shown an inverse growth trend.Cognitive radio technology is one of the effective ways to solve the shortage of spectrum resources,in which spectrum sensing technology is the primary technology of cognitive radio.At present,the test results of the more mature single-mode spectrum sensing technology cannot fully reflect the information of primary users,which has certain limitations in practical applications.Multi-modal spectrum sensing technology has received widespread attention as an effective solution.This thesis takes cognitive radio as the research background and focuses on the spectrum sensing technology.On the basis of summarizing and analyzing the existing spectrum sensing technology,with the help of D-S evidence theory to carry out in-depth research on multi-modal sensing technology,the main work is as follows:1.The existing multi-modal spectrum sensing D-S theory fusion rules adopt classic conflict coefficients to characterize the total degree of conflict of global detectors,ignoring conflicts among detectors,which lead to an inability to accurately describe the trust of each detector,thereby affecting the final decision result.Aiming at this problem,this thesis proposes a multi-modal cooperative spectrum sensing algorithm based on improved D-S evidence theory,which uses extended conflict coefficients to characterize the degree of conflict between the detection results of each detector and the detection results of other detectors.The fusion center groups the detection results of multi-modal detectors and calculate the basic credibility function of each detector combination,probability conversion,calculation of conflict coefficients,discount conflict coefficients,and improved fusion rules,so that the multi-modal detection result can be accurately measured,the fusion coefficient is reduced in time during the fusion process,the conflict between the detection evidences of the multi-modal detector is reduced.The detection performance of multi-modal cooperative spectrum sensing is improved.2.For the existing multi-mode cooperative spectrum sensing,the correlation between the multi-mode detection data on the secondary user(SU)is ignored.When the number of SUs increases,the amount of data transmitted on the report channel increases rapidly,resulting in data redundancy and greater system overhead.This thesis proposes a step-by-step multi-modal fusion spectrum sensing method.Firstly,on the secondary users,the Bayesian method was used to analyze the relationship between the detector data of different detectors,and the weighted sum method was used to realize the local decision,which reduced the amount of data transmitted to the fusion center.Secondly,the fusion center determines the credibility of each secondary user based on the historical data of the decision result,and uses D-S evidence theory.The theoretical fusion results in a global decision to ensure the accuracy of spectrum sensing.The final simulation results show that the detection accuracy of the proposed algorithm is not much different from the existing multi-modal spectrum sensing,and the amount of information transmitted in the data is reduced to about one-third.