宽带认知无线电网络分布式协作压缩频谱感知算法  

A Sensing Algorithm Based on Distributed Cooperative Compressed Spectrum for Wideband Cognitive Radio Networks

作　　者：张正浩[1] 裴昌幸[1] 陈南[1] 李虎生[2] 杨德鹏[2] 

机构地区：[1]西安电子科技大学综合业务网国家重点实验室,西安710071 [2]田纳西大学电子工程与计算机科学系,美国诺克斯维尔37996

出　　处：《西安交通大学学报》2011年第4期67-71,114,共6页Journal of Xi'an Jiaotong University

基　　金：国家自然科学基金资助项目(61072067;60970119);美国国家自然科学基金资助项目(CCF-0830451);高等学校创新引智计划资助项目(B08038);国家重点实验室专项资金资助项目(ISN1001004);陕西省工业攻关计划资助项目(2009K01-46)

摘　　要：针对宽带认知无线电网络中压缩频谱感知算法在低信噪比环境下频谱检测性能下降的问题,提出了一种基于高斯过程的分布式压缩频谱感知(PBCS)算法.首先应用层次化的正态分布概率模型来表示压缩频谱的重构,然后各个认知无线电用户交换模型参数并结合本地的压缩采样数据进行压缩频谱感知.有别于其他直接融合频谱感知结果或检测数据的协作式算法,PBCS算法通过模型参数融合来进行协作,能有效减小信噪比低的协作用户的影响,从而提高算法的抗噪性.仿真结果表明,PBCS算法可以在-5dB的信噪比条件下达到检测概率大于0.9、误检概率为0.1的频谱检测性能.A distributed probabilistic Bayesian compressed spectrum sensing algorithm （PBCS） is proposed based on Gaussian process to improve the detection performance of compressed spectrum sensing under low signal to noise ratio condition in wideband cognitive radio networks. A hierarchical normal distribution probabilistic model is applied to represent the compressed spectrum reconstruction. The model parameters are exchanged among the cognitive radios as cooperation information, and then are utilized to implement compressed spectrum sensing based on the local compressed sensing data. Being different from existing cooperative spectrum sensing methods, where the final sensing decisions or the rough sensing data are used as the cooperation information, model parameters fusion is conveyed among the cognitive radios in the proposed PBCS algorithm. Therefore, the negative effectiveness from those cooperated neighbors experiencing low signal to noise ratio sensing environment is effectively reduced and the detection performance is improved. Simulations show that PBCS achieves a detection rate over 0. 9 while false alarm remains 0. 1 when signal to noise ratio is -5 dB.

关 键 词：宽带认知无线电网络 高斯过程 分布式压缩频谱感知 模型参数融合 

分 类 号：TN929.5[电子电信—通信与信息系统]