Research On Cooperative Spectrum Sensing Technologies Based On DS Evidence Theory

In modern information battlefield,the key to win the information war is to obtain the right of information first in the complex electromagnetic space,and the first task is to carry on the full-dimensional situation sensing to the battlefield electromagnetic spectrum.Cooperative spectrum sensing technology can obtain the electromagnetic situation of the battlefield,so as to make accurate judgment and effective prediction of the battlefield situation.However,in the complex electromagnetic space,the sensing information is affected by the channel environment and malicious attacks and is uncertain,which leads to inaccurate fusion results.Therefore,DS evidence fusion is applied to cooperative sensing to improve the reliability of fusion results.However,due to the adverse electromagnetic environment and the drawbacks of the DS combination formula,the traditional sensing methods based on DS fusion have the problem of evidence paradox,which seriously destroys the accuracy of the sensing results.Therefore,in the face of this severe challenge,it is necessary to find a reasonable solution to ensure the accuracy and reliability of the sensing results on the basis of improving the detection performance of spectrum sensing.The research goal of this article is the problem of evidence paradox.From the point of view of modifying evidence values and improving combination formula,two different solutions are proposed,which effectively solve the problem of evidence paradox and improve the reliability of sensing results.The main research contents are summarized as follows:Firstly,the system model of cooperative spectrum sensing is established,and DS evidence theory is applied to fuse sensing information.In view of evidence paradox in this method,the causes and classification of evidence are analyzed,and the solutions are summarized into two categories: modifying evidence values and improving combination formula.Secondly,based on the idea of modifying the evidence values,the sensing method based on the similarity between evidences is proposed.In this algorithm,a weighted distance measure is constructed to reflect the degree of similarity between sensing nodes,and the degree of similarity is converted into weights to modify the sensing information.The feasibility of the distance measure is verified by theoretical derivation,and the sensing performance and robust performance of the evidence similarity method are verified by simulation analysis.Finally,based on the idea of improving the combination formula,the sensing method of evidence preprocessing is proposed.In this algorithm,static weights and dynamic weights are used to preprocess the sensing information,and then a combination formula based on evidence preprocessing is proposed to fuse the preprocessed sensing information.A combination formula based on evidence preprocessing is proposed to fuse the sensing information after preprocessing.By comparing the simulation analysis with other scholars' methods,the sensing performance and robust performance of the evidence preprocessing method are verified respectively.