Research On Spectrum Detection Method For Cognitive Radio Networks Based On Signal To Noise Ratio Selection And Weighting

With a continuous increase of demand quantity for the radio network service, the relevant volume of radio communication has difficultly met the needs of people, because of the changeless spectrum allocation policy adopted by the traditional radio network. Meanwhile, the utilization rate of the frequency band assigned to licensed user is very low. With the presentation of cognitive radio technology, the contradiction on utilizing spectrum resource between unlicensed user(cognitive user) and licensed user(primary user) can be effectively relieved. Spectrum detection is an important and crucial link for the cognitive radio technology, and it has been a research emphasis for the scholars and experts. In this paper, the spectrum detection technology was deeply studied, and the multi-user collaborative spectrum detection method was a research emphasis.Firstly, the development background, research situation and application significance of the cognitive radio technology was studied. The contrast research on several kinds of typical spectrum detection algorithm for the single user was carried out. Specific research contents included energy detection, matched filter detection, cyclostationary feature detection, interference temperature model detection and local oscillator power leakage detection. Thereinto, the research on energy detection algorithm was thorough. Further research on collaborative spectrum detection method was performed, aiming at overcoming the limitation of single node spectrum detection algorithm. Detailed contents included the collaborative spectrum detection technology and relevant data fusion rules.Secondly, the targeted research contraposing an important issue in the cognitive radio system was conducted, that is the vacant spectrum in frequency band licensed by the master user in the cognitive radio network could be efficiently utilized. When the number of licensed user was very large, the detection effect was non-ideal, though the traditional data fusion collaborative spectrum detection algorithm was simple and easy to implement. In order to further improve the detection performance, a multi-user data fusion collaborative spectrum detection algorithm based on signal to noise ratio selection was presented, and the simulation test for the presented method was carried out. The simulation results showed that the detection probability could be improved and the detection reliability could be enhanced to some extent, compared with the traditional data fusion collaborative spectrum detection algorithm.Finally, a grouping collaborative spectrum detection scheme based on signal to noise ratio weighting double threshold was presented, which contraposed the problems of the excessive communication request quantity on the control channel, the loss of bandwidth resources and the unreliable result of judgment. This scheme combined signal to noise ratio weighting double threshold detection with grouping collaborative detection. All the cognitive users were grouped, and one cluster-head was filtrated from every group. The master user was firstly detected by the intragroup cognitive user, then the result of judgment or detection energy value was sent to the cluster-head according to signal to noise ratio weighting double threshold detection method. The cluster-head applied again it to detect, and the result of judgment was achieved and reported to the fusion center. The center used OR fusion to blend all results, then the final judgment was obtained. Different weights were endowed by signal to noise ratio weighting double threshold detection method based on the size of signal to noise ratio for the cognitive nodes, and the effect of size of weight on fusion judgment could be controlled. The simulation results showed that the spectrum detection probability could be effectively improved, the communication request on the control channel could be reduced and the loss of system resource could be decreased by using the presented method.