Research On Spectrum Sensing Of Double Secondary Users In Cognitive Networks

To do spectrum sensing in reality,the received signals mostly contain unknown parameters.In addition,the traditional spectrum sensing algorithms cannot reach the performance requirements.For making the primary user(PU)accesses the licensed spectrum again less passively and not limited to focus on one secondary user(SU),this paper studies a radio cognitive system with one PU and two SUs.First,the PU detects the signal transmitted in the noise to determine whether SUs exist.The estimation of unknown noise variance is calculated based on maximum posteriori probability(MAP)criterion with known prior distribution.The final detection result can be obtained by GLRT-based spectrum sensing algorithm.Furthermore,considering the actual channel environment,the transmitted signals from two SUs have a delay and the integral operation is used to eliminate its impact.Secondly,the PU determines whether one SU or two SUs to have access to the frequency band according to the received signals.Detection algorithm based on GLRT with MAP estimation is also used to obtain the test results here.At the same time,taking into account the possible influence on the test results in different prior distribution,Inverse-chi distribution,inverse Gaussian distribution and non-informative distribution are used as priori knowledge to obtain different detection results.The simulation performance of the algorithm proposed in this paper based on MATLAB platform makes the probability of detection as a standard.At the same time,this paper also discusses the influence of different prior distributions to the detection performance.Simulation results show that increase of the number of samples and false alarm probability both can improve the detection probability.In addition,compared with spectrum algorithm based on GLRT with ML estimation,the proposed algorithm in this paper reduces the number of samples significantly and gets good performance.