【摘要】 认知无线电(CR)被提出来是为了解决频谱资源短缺和低频谱利用率的问题。它允许认知用户在授权用户空闲的时候机会地接入未占用的频谱。因此,快速而又准确地检测授权用户信号是认知无线电系统的基础。为了提高认知无线电网络协作频谱检测的性能,利用了最优化的本地频谱检测,并结合了基于簇的网络模型。首先,将认知用户分为若干个簇,并选出簇头节点。在同一个簇中的认知用户,共同做出一个簇判决。之后,各个簇头节点进行通信做出全局判决。此外,信噪比(SNR)高的认知用户,向信噪比低的认知用户发送辅助信息,以帮助它们利用自适应双门限法,更加准确地完成本地频谱检测。通过仿真可以发现,本地频谱检测和全局频谱检测的性能都得到了提高。更多还原

【Abstract】 Cognitive Radio(CR) has been proposed to deal with the problems of scarce spectrum resources and low spectrum utilization which makes the CR users access to the idle spectrum band when the licensed user is absent.Thus,detecting the licensed user’s signal quickly and accurately is the foundation of the CR system.For improving the performance of the cooperative spectrum sensing in the CR network,the optimized local spectrum sensing and the network model based on cluster have been utilized.Firstly,the CR users are divided into some clusters and chosen the cluster-head nodes.The CR users in the same cluster make a cluster decision together,and then the cluster-head nodes conduct a communication to make the global sensing decision.Besides,the CR users with high SNR will send assistant information to the ones with low SNR in order to help them complete the local spectrum sensing by double adaptive thresholds energy detection.Through the simulation not only the local spectrum sensing’s performance but also the global cooperative spectrum sensing’s performance has got improved.更多还原