The Study On The Technology Of Spectrum Sensing In Cognitive Radio System

Wireless spectrum is a kind of rare nature resource and its distribution and utilization are usually regulated by radio management departments. Today the management method of wireless spectrum by using all over the world limits the effectiveness of the use of the spectrum, which usually makes part of the spectrum unused. In this case it can't meet the growing demand of wireless communication services. Cognitive Radio (CR) can change the sending and receiving parameters adaptively according to its interactive sensing of the wireless communication condition it works in. Then Cognitive Radio can use the available spectrum repeatedly, which improves the utilization of the available spectrum effectively to avoid interfering the primary user.The research contents of this paper are as follows:Firstly it gives a brief introduction to IEEE 802.22 WRAN, the emergence background and research model of CR, and its studies status at home and abroad. Then we briefly introduce the application trends of Cognitive Radio system in the chapter. Chapter Two starts a general discussion about the spectrum detection techniques in the Cognitive Radio system. And their advantages and disadvantages are concluded.The basic theory of cyclostationarity is introduced in the third chapter. A new spectral estimation algorithm is put forward, which improves the performance and the complexity of spectral estimation by method of average. In addition, the cyclostationary of the general communication signal is analyzed. And it further studies the cyclostationary of the fading signal and proves it on the computer by simulation of the fading to the cyclostationary and Doppler frequency shift to the cycle frequency. All these provide a theoretical basis for use of cyclostationary feature detection in cognitive radio. Results of simulation confirm their performance.The fourth chapter is mainly about the use of cyclostationary feature detection in cognitive radio. A constant false alarm rate (CFAR) multi-cycle detector based on a chi-squared statistical test is used in the spectrum sensing of cognitive radio. In order to overcome the shortcoming that the detection based on one cycle frequency can't make full use of the information contained in signals, the paper proposes a new method based on multi-cycle frequency, which make full use of the information and improve the reliability of the signal detection. Considering that the single detection can't detect the first signal in shadow or depth decline, another new collaboration detection based on a number of cognitive notes is proposed and the chapter analyzes the advantages of this method to the single detection. Finally simulation results prove its effectiveness.