Research On The Cyclostationary Spectrum Sensing In Cognitive Radio

With the rapid development of wireless communication technology, thecontradiction between the growing demand of spectrum and the limit of spectrumresource has become sharper and sharper. In order to resolve this issue, there emergedCognitive Radio technology. Hence, spectrum sensing has received wide attention andbecome more and more popular in recent years as it is the basis of the cognitive radiotechnology.In this paper, we concentrate our attention to relatively deep study of spectrumsensing technology, especially focused on the cyclostationary detection algorithm,which involves two aspects. One is the study of the local cyclostationary algorithm andthe other is the study of the combination of the cyclostationary and collaborativedetection in order to improve the local cyclostationary algorithm.When it comes to terms with the local cyclostationary detection, the traditionalcyclostationary detection algorithm needs to be improved as it has several disadvantages,such as the large computational complexity and the difficulties in the realization of realtime detection. Here we proposed an improved one named cyclostationary detectionalgorithm, by building the cyclic spectrum, which was simplified in the frequencydomain and thus can reduce the complexity greatly even when we get very little prioriknowledge before the detection. To facilitate the use of Matlab functions, this papersupplemented the expression for the formula of false alarm probability. Additionally, adirect relation between the detection probability and the signal-to-noise ratio in theactual wireless environment is obtained.In the collaborative cyclostationary detection, the local detection has the problemof the uncertainly of authorized receiver’s position and hidden terminal. So，in thisresearch, we combined the cyclostationary and collaborative detection to overcome thedrawbacks of collaborative cyclostationary detection. In the hard decision, we found outthat the fusion center uses traditional ‘and’ rule and ‘or’ rule, respectively, to dosimulation analysis in the different scenarios. While, in the soft data fusioncollaboration detection, the fusion center adopts equal values merger algorithm, whichwas particularly analyzed in this paper comparing with the hard decision fusioncollaboration in simulation.Through the analysis and comparison, it turns out that the collaboration cyclostationary algorithm based on equal values is superior to that based on the harddecision and the soft data fusion collaboration detection has much better performance,however, they still cannot replace the hard decision fusion collaboration detectioncompletely. The simulation turns out that when the interference limit is not rigid,‘or’rule can almost reach the detection performance of the soft data fusion collaboration.Therefore, we should base on actual requirements when choosing the type ofcollaborative detection.