Research On Spectrum Sensing Algorithms Based On Goodness-of-fit Test

With the rapid development of the times,wireless communication technology has greatly facilitated the daily life of the public.However,at the same time,the problem of scarcity of spectrum resources is getting more and more serious due to the ever-increasing demand for spectrum resources and the fixed spectrum allocation strategy.Cognitive radio is an emerging technology to improve the utilization of spectrum resources.It allows cognitive users to rationally reuse the spectrum of primary users without interfering with the communication of primary users.Efficient spectrum sensing technology is the basis and premise of cognitive radio technology,and has received widespread attention and research.The spectrum sensing algorithm in cognitive radio technology is the core research object of this paper.In many perceptual algorithms,the algorithm based on goodness of fit will be focused on because of its excellent characteristics.Even if the algorithm is unknown to authorized users Information can also achieve excellent detection performance with a small number of received signals.However,existing GOF sensing algorithms can only perceive static signals with high efficiency,and the performance declines when sensing non-static signals.For this reason,this paper presents a new goodness-of-fit detection algorithm using the distribution characteristics of the largest eigenvalues.The new algorithm combines the knowledge of the stochastic matrix theory to calculate new fitting statistics based on the maximum eigenvalues,so that efficient detection of authorized users can be achieved when detecting various signals.In addition,when analyzing the distribution rule of the fitting object in the goodness-of-fit test algorithm,it is found that the value of the partially-fitted object is always larger than that of the authorized user when there is an authorized user.Based on this idea,this paper designs a new unilateral-right-tailed AD fitting criterion,which not only reduces the computational complexity of the goodness-of-fit test algorithm,but also further improves the detection performance.The new detection algorithm and fitting criteria proposed in this paper provide the possibility to solve some problems in the goodness-of-fit test spectrum sensing technology.