基于压缩感知的认知无线网络协作频谱感知技术

Novel Cooperative Spectrum Sensing Method Based on Compressed Sensing for Cognitive Radio Networks

针对传统频谱感知算法受A/D器件物理瓶颈,硬件计算能力等限制难以实现宽频段授权用户实时检测的问题,将压缩感知技术应用于宽带频谱感知,建立了一个基于Xampling前端的新型协作频谱感知模型,在此模型基础上提出一种基于认知用户分组,组内利用压缩感知技术进行信息融合,组间对各组检测结果进行判决融合,两者结合实现授权用户检测的新算法,并给出了认知用户的分组原则。与传统宽带频谱感知算法相比,该方法无需高速A/D,也不需要恢复原始信号,可有效降低宽带频谱感知的硬件复杂度,缩短频谱感知时间,仿真结果证实了新算法的有效性。

Limited by the physical bottleneck of A/D device,hardware computing power and other restrictions,traditional spectrum sensing methods can hardly realize the real-time detection of the licensed users in wide-band spectrum.To overcome this problem,compressed sensing was invited into wide-band spectrum sensing,and a new collaborative wide-band spectrum sensing model was established based on Xampling.A new algorithm was proposed based on cognitive user groups,in which information fusion based on compressed sensing and decision fusion were combined.Compared with the traditional sensing algorithms,this method needs either high-speed A/D device or reconstruction of the original signal,which can effectively reduce the complexity of wide-band spectrum sensing hardware and the sensing time.Finally,simulation results show the validity of the proposed scheme.