The Research Of Efficient And Robust Spectrum Sensing Algorithm In Cognitive Radio System

Recent years, the type of wireless business increase quickly but the spectrum resources are scarce relatively. In order to use existing spectrum resources efficiently,the concept of cognitive radio arises at the historic moment. When the primary users doesn’t use the spectrum the other users also can’t use it. So it will cause a huge waste of spectrum resources. Spectrum sensing is used to detect such "spectrum holes" and take advantage of them and it is a important technology of cognitive radio.To the actual application, spectrum sensing technology should be efficient and robust in order to guarantee the stability of cognitive radio network. Because of the existing algorithms based on the eigenvector involves a large number of eigenvalue decomposition, so it didn’t fit for the real-time detection. The proposed algorithm utilizes the orthogonality between the signal subspace and the noise subspace and projects the received signal of the secondary users on the two subspace to achieved fast spectrum sensing based on the differences of the projection values.The theoretical analysis and the simulation results show that the proposed algorithm reduces the computational complexity effectively and it is robust to the noise uncertainty.Then in view of the performance of the existing spectrum sensing algorithm based on goodness of fit testing becomes worse when there is noise uncertainty, this paper employs the moment estimation method or the eigendecomposition estimation method to estimate the noise variance in time, and using the estimated value, lowers the complexity of the algorithm and avoids the influence of noise uncertainty by processing the sampling data appropriately. Simulation results show the effectiveness of the proposed method in alleviating the noise uncertainty.Anyhow, spectrum sensing technology can meet the growing demand of the spectrum resources and ensure the quality of communication.It provides a new way to improve resource utilization and guarantee the reasonably use of resources. So the research of spectrum sensing is necessary and urgent, and how to make the spectrum sensing more efficient and more robust becomes the top priority of cognitive radio technology.