基于压缩感知信道能量观测的协作频谱感知算法  

Cooperative Wideband Spectrum Sensing Algorithm Based on Compressed Sensing Channel Energy Measurements

作　　者：顾彬[1] 杨震[1] 胡海峰[1] 

机构地区：[1]南京邮电大学"宽带无线通信与传感网技术"教育部重点实验室,南京210003

出　　处：《电子与信息学报》2012年第1期14-19,共6页Journal of Electronics & Information Technology

基　　金：国家973计划项目(2011CB302903);国家自然科学基金(60971129;61071092);江苏省"信息与通信工程"优势学科基金(YX002001)资助课题

摘　　要：压缩感知为认知无线电宽带频谱感知提供了一种新思路。基于压缩感知原理,该文提出一种不需要重构宽带频谱本身,而是直接重构各信道能量的协作频谱感知方法。多个次用户使用宽带随机滤波器组获取信道能量的观测值。融合中心同步接收多个用户的能量观测,并利用同步稀疏自适应匹配追踪协作重构算法重构所有次用户的信道能量。仿真结果表明加性高斯白噪声环境下该协作感知方法所需的滤波器数目仅为传统方法的20%左右,瑞利衰落信道下也仅需传统方法的40%,有效降低了系统复杂度并改善感知性能。同时,该文提出的同步稀疏自适应匹配追踪算法对比经典的同步正交匹配追踪算法在重构精度及算法复杂度两方面都有所提升。Compressed sensing offers a new wideband spectrum sensing scheme in cognitive radio.This paper presents a cooperative sensing scheme based on compressed sensing to sense channel energies without reconstructing the wideband spectrum.Multiple secondary users employ a number of wideband random filters to achieve channel energy measurements.A centralized fusion center is used to collect simultaneously the measurements where a novel cooperative recovery algorithm named Simultaneous Sparsity Adaptive Matching Pursuit（SSAMP） is utilized to reconstruct all the channel energies.Simulations show that the cooperative scheme only needs 20% of the required number of filters in additive white Gaussian noise channel and needs 40% in Raleigh fading channel.SSAMP algorithm outperforms the Simultaneous Orthogonal Matching Pursuit（SOMP） on both reconstruction quality and algorithm complexity.

关 键 词：认知无线电 宽带频谱感知 压缩感知 能量检测 

分 类 号：TN92[电子电信—通信与信息系统]