Research On Spectrum Sensing Technology In Cognitive Radio Network

The main purpose of spectrum sensing is to find free spectrum resources.Traditional spectrum sensing research has some problems,such as: the number of cooperative users is small,and the performance of low signal-to-noise ratio is not ideal;the calculation process of the decision threshold is complicated and the result is inaccurate;the sensing data there is abnormal data interference.In order to avoid the above problems,this paper proposes two cluster-based spectrum sensing algorithms,the specific content is as follows:In order to increase the number of collaborative users and extract signal features more accurately,a feature extraction method based on IQ decomposition and Covariance Absolute Value(CAV)is proposed.In order to avoid deriving the threshold value,a spectrum sensing algorithm based on DBSCAN(Density-based spatial clustering of applications with noise)clustering is proposed.This method first extracts the signal features of the sensing signal to form a two-dimensional feature vector,then uses the feature vector to train the DBSCAN clustering algorithm to obtain the corresponding classifier,and finally uses the classifier to determine whether there is a primary user and realize spectrum sensing.In order to reduce the effect of abnormal sensing data sent by Secondary User(SU)on the performance of the spectrum sensing system,a robust cooperative spectrum sensing algorithm based on information geometry is proposed.The algorithm is divided into two stages: preprocessing and data fusion.In the preprocessing stage,the data is decomposed by IQ to obtain two sets of data.Then select the reference point on the manifold,FC calculates the Riemann distance between all SUs and the reference point;in the data fusion stage,the average value of each group of all covariance matrices and the reference point Riemann distance is used as a statistic,and finally constitutes a twodimensional signal Feature vector.Finally,the Fuzzy c-means clustering algorithm is trained to obtain a classifier,and then the classifier is used to realize spectrum sensing and realize spectrum sensing.