基于非参量累积和的合作频谱感知

针对认知无线电网络中频谱感知的检测时延降低问题，提出了基于非参量累积和的合作频谱感知算法。本地认知用户预处理频谱观测数据，获得观测数据相对于信念值的正向漂移和负向漂移。为了缩短检测延迟，认知用户只将数据的正向漂移同步传输至融合中心。融合中心融合正向漂移得到判决信息，采用非参量累积和算法依时间序列顺序累加判决信息，判断主用户是否正在使用授权频段。为了解决不传输负向漂移引起的虚警问题，改进算法提出融合中心可以保留首次判决，经过等待时间间隔后再作出最终判决。相对于传统的软融合算法，改进融合规则的合作频谱感知算法具有较低的检测延迟。

Cooperative spectrum sensing based on nonparametric cumulative sum

For the problem that the detection delay of spectrum sensing is decreased in the cognitive radio(CR) network,a cooperative spectrum sensing method based on nonparametric cumulative sum(CUSUM) is proposed.Local CR users preprocess the observation data of spectrum to obtain positive drifts and negative drifts relative to the reference value.To reduce the detection delay,CR users only transmit positive drifts to fusion center.Fusion center fuses positive drifts to achieve decision information and decide whether primary users are using the authorized spectrum by adopting nonparametric CUSUM to accumulate decision information.To settle the false alarm problem caused by not transmitting negative drifts,the proposed method notes that fusion center can hold the initial decision until latency time interval is passed and make the final decision.The proposed method with new fusion rule has better performance in reducing detection delay with respect to conventional soft combination scheme.