Spectrum Sensing In Cognitive Radios Networks

Nowadays,under the huge wireless communication environment,the number of Internet users has sharply increased,the use of spectrum resources become more strained.On one hand,each country will set aside some of the band open to industrial, scientific and medical institutions, which are known as the ISM band. The ISM band is freely licensed,and it can be directly used if there’s no influence to other bands. This situation will cause spectrum congestion. On the other hand, many bands that do not have a license which is not reasonablely used and some hardware equipments are not up to the requirements of using high frequency.For the above two situations,Dr.JosephMitola from Sweden put forward to the concept of cognitive radio,American FCC set working group to carry on the research of cognitive radio in 2002.Based on the development of cognitive radio,some companies which led by Motorola put forward the technology of cognitive radio network. The appearing of the cognitive radio network can effectively improve the utilization rate of spectrum resources. The research of cognitive radio networks key technology mainly includes two aspects: On one hand is the spectrum sensing technology,On the other hand is the dynamic allocation of spectrum.This paper focuses on the spectrum sensing. Main achievements are following:Fistly,Preprocessing of signal,namely removal the noise of sense band,The useful signal will be separated from the noise, which can improve the signal-to-noise ratio. Single user ofen uses Matching filter, Cycle characteristics and Energy detection algorithm to detect respectively the interested frequency bands.The simulation conclusion shows that under the condition of high SNR, energy detection has better effect.Secondly, the single node is easily affected in multipath and shadow, so it can not accurately judge the state of primary users. To solve the shortage, this paper proposes cooperative sensing algorithm for multiple users. There is a fusion center, in which mutual sharing between cognitive users is achieved, then the sensing result is sent to the fusion center. According to the received information the fusion center conduct AND, OR rule fusion,then the verdict is got.But under the uncertain noise, according to the energy detected by multiple users, the traditional single threshold value is not easy to determine, which can cause prone error. So this paper puts forward the double threshold energy detection algorithm, because it can increase the uncertainty region of a threshold threshold value. The experimental results show that, the double threshold is more accurate than the traditional single threshold detection. But the double threshold energy detection will only detect energy value falling on both ends of the threshold, and the value falling in the middle will not be detected.Finally, the energy value between the double threshold will be detected by the by the covariance detection algorithm, and the traditional covariance detection algorithm uses fixed threshold and false alarm probability, so it cannot effectively make the perception error reach a minimum in any cases. Therefore, according to this shortcoming, traditional covariance threshold is improved on the threshold value in this paper. It can be proved that the proposed method can accurately sense the idle spectrum resources through the experiment simulation.