低信噪比下基于双检测长度的频谱感知算法  

Spectrum sensing method based on the double sensing length under a low SNR

作　　者：高锐[1] 李赞[1] 齐佩汉[1] 关磊[1] 刘进[1] 

机构地区：[1]西安电子科技大学综合业务网理论及关键技术国家重点实验室,陕西西安710071

出　　处：《西安电子科技大学学报》2014年第3期88-94,共7页Journal of Xidian University

基　　金：国家自然科学基金资助项目(61072070);教育部博士学科点基金资助项目(20110203110011);教育部基础科研业务费资助项目(72124338);陕西省自然基金重点资助项目(2012JZ8002);ISN国家重点实验室自主课题资助项目(ISN1101002);高等学校学科创新引智计划资助项目(B08038)

摘　　要：针对现有认知网络中低信噪比条件下频谱感知性能差、速度慢的问题,提出了一种基于双检测长度的感知算法.该算法先进行一次短数据长度能量检测,当第1次判决主用户不存在时,再进行第2次长数据长度能量检测,并作出最终的感知判决.从理论上推导了平均样本数、检测性能、算法复杂度等重要性质.基于matlab的仿真结果验证了理论推导的正确性,表明所提算法在低信噪比条件下,在感知性能较传统能量检测算法有所提升的同时,具有更少的平均样本数和更低算法复杂度的特点,并有望应用于未来实际认知网络中.According to the spectrum sensing problem under a low signal-noise ratio （ SNR） in cognitive radio networks , this paper proposes a spectrum sensing method based on the double sensing length . For the proposed algorithm , a short length energy detection will be carried out first . While it is making the decision there is no primary user , and the second long length energy detection will be taken to make the final decision . This paper analyzes some significant properties , such as the average sample number （ ASN ） , detection performance and implementation complexity . Simulation results validate the theory , and show that in the low SNR regime the proposed algorithm can make the spectrum sensing more efficient than conventional energy detection . Meanwhile it possesses the advantages of a smaller ASN and lower complexity , and has potential applicability in cognitive radio sensor networks .

关 键 词：认知无线电 频谱感知 能量检测 双检测长度 

分 类 号：TN911.72[电子电信—通信与信息系统]