基于ATmega128的连续信号压缩感知实现方法  

Method for Continuous Signal Compressed Sensing Based on Atmega128

作　　者：王海安 郭改枝 张鹏举 WANG Haian;GUO Gaizhi;ZHANG Pengju(School of Computer Science and Technology,Inner Mongolia Normal University,Hohhot 010020,China)

机构地区：[1]内蒙古师范大学计算机科学技术学院,呼和浩特010020

出　　处：《内蒙古大学学报（自然科学版）》2023年第2期206-212,共7页Journal of Inner Mongolia University：Natural Science Edition

基　　金：内蒙古自然科学基金项目(2020MG06029,2021LHMS06013,2020LH06009);内蒙古自治区科技计划项目(2020GG0165);内蒙古高等学校科学研究项目(NJZY21553);内蒙古师范大学基本科研业务费专项资金资助(2022JBXC018);内蒙古师范大学研究生科研创新基金资助项目(CXJJS22141)。

摘　　要：为了解决硬件对高频模拟信号采集困难的问题,根据连续信号可以在特定的频域有稀疏的特性,在采样节点,根据伯努利观测矩阵分布特点设计采样间隔,利用压缩采样方法采集数据,降低收发芯片CC1000所需发送的数据量,从而降低传感器节点在通信传输时所损耗的能量,延长节点的使用时间;在汇聚节点使用压缩采样匹配追踪算法(Compressive Sampling Matching Pursuit,CoSaMP)对接收到的稀疏数据进行重构,将重构的数据通过串口助手导出到终端,使用Matlab软件绘图分析,验证压缩感知实现方法的可行性。实验结果表明:该方法能够很好的在ATmega128芯片上实现压缩观测。在压缩率达到68%时,通过分析计算原始信号波形和重构信号波形,二者具有99.30%的相似度,显然将压缩感知理论可应用到局域网中,能减少网络传输的数据量,降低网络传输的能耗。In order to solve the difficulty of hardware acquisition of high frequency analog signal,according to the continuous signal can be sparse in the specific frequency domain,the sampling interval is designed according to the distribution characteristics of Bernoulli observation matrix at the sampling node,and the compression sampling method is used to collect data to reduce the amount of data needed to be sent by the transceiver chip CC1000,so as to reduce the energy lost by the sensor node in the communication transmission and extend the service time of the node;the Compressive Sampling Matching Pursuit(CoSaMP)is used at the sink node to reconstruct the received sparse data and export the reconstructed data to the terminal through the serial port assistant.The reconstructed data is exported to the terminal through the serial assistant,and the Matlab software is used for drawing and analysis to verify the feasibility of the compressed sensing method.The experimental results show that compression observation on the ATmega128 chip can be well achieved by this method.When the compression rate reaches 68%,by analyzing and calculating the original signal waveform and the reconstructed signal waveform,the two have a 99.30%similarity.Obviously,the compression sensing theory can be applied to the LAN,which can reduce the amount of data transmitted by the network and reduce the energy consumption of network transmission.

关 键 词：压缩感知 伯努利观测矩阵 压缩采样匹配追踪算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]