基于预测的差分信号压缩感知算法  

Differential Signal Compressed Sensing Algorithm based on Prediction

作　　者：王盼盼[1,2] 季薇[1,2] 郑宝玉[1,2] 

机构地区：[1]南京邮电大学通信与信息工程学院,江苏南京210003 [2]南京邮电大学教育部"宽带无线通信与传感网技术"重点实验室,江苏南京210003

出　　处：《信号处理》2015年第11期1446-1453,共8页Journal of Signal Processing

基　　金：国家863项目(2009AA01Z241);江苏省自然科学基金项目(BK20140885);江苏省博士后科研资助计划(1401045C);南京邮电大学科研基金项目(NY214034);国家自然科学基金项目(61471200)

摘　　要：基于压缩感知的频谱感知方法可以较低的采样速率快速获取信号,并利用获得的稀疏数据样本来判断信道的占用情况。然而,压缩感知技术中信号重构算法的复杂度很高,难以满足无线通信中的实时性要求。本文提出一种基于预测的差分信号压缩感知算法,该算法利用信道占用时间上的相关性,建立了一种信道占用情况的预测模型,依此模型预测出信道占用的变化情况;基于预测结果,在重构信号时可减少频点的搜索范围,两次降低重构算法的运算量。仿真结果表明,在保证感知性能的前提下,新算法可大幅降低迭代次数,减少算法复杂度。Compared with traditional spectrum sensing methods,compressed sensing based spectrum sensing can fast obtain a signal with a lower sampling rate and then use the obtained data to determine the channel occupancy. However,the reconstruction algorithm in compressed sensing techniques is usually too complex to meet the requirements of real-time communication. A prediction based differential signal compression sensing algorithm is proposed in this paper,which utilize time correlation of the channel to build a prediction model and predict the changes of channel occupancy. In following process of reconstruction,based on the prediction result,the search range of frequency points can been reduced,thus reducing the number of iterations of the reconstruction algorithm. Simulation results show that the new algorithm can significantly reduce computational complexity and ensure good sensing performance at the same time.

关 键 词：压缩感知 频谱感知 正交匹配追踪 差分信号压缩感知 

分 类 号：TN929.5[电子电信—通信与信息系统]