压缩感知算法的改进及其在无线传感网络中的应用

发布时间：2018-06-15 17:03

本文选题：无线传感网络 + 压缩感知　；参考：《燕山大学》2016年硕士论文

【摘要】：无线传感器网络综合了传感、无线通信及嵌入式系统等相关技术,可在多种场景下,通过传感器节点终端进行数据采集和处理,获取大量可靠详实的信息。压缩感知理论能够将稀疏或者可压缩信号中的“冗余”信息丢弃,降低信号懫样频率,节省存储和传输成本,为无线传感器网络的通信、数据采集、定位等问题提供了一种全新的技术解决方案。本文主要研究压缩感知技术在虚拟MIMO无线传感网络中的自适应通信与定位等应用问题,从而改善无线传感网络的通信性能,提高节点定位的精度。首先对压缩感知理论进行了概述,并对几种常用的压缩感知重构算法进行了详细阐述。针对传统压缩感知算法复杂度高、收敛速度慢、重构精度低等不足,在压缩感知基本理论基础上,利用陡峭性更好的近似双曲正切函数来优化平滑0l范数重构算法(Smoothed 0l,SL0)的逼近函数,同时根据混合优化算法来进行寻优,从而提出一种改进的SL0压缩感知算法。其次将改进的压缩感知重构算法应用于虚拟MIMO无线传感网络自适应通信信道状态信息反馈中,降低信道状态信息反馈量,减小信道开销,且能够在减小反馈率的同时,增加系统最大总速率以及反馈信息的准确性,进而提高网络性能,为无线传感网络定位提供良好的通信保障。最后,在保证无线传感网络通信系统实现准确通信的前提下,将改进后的压缩感知算法与RSS定位技术相结合,通过RSS构建压缩感知传感矩阵,运用改进的算法进行信号重构,提高定位的实时性、准确性与抗噪声性,实现高精度、快速无线传感网络多目标定位。
[Abstract]:Wireless sensor network (WSN) integrates sensor, wireless communication and embedded system technologies. It can acquire a large amount of reliable and detailed information through sensor node terminal data acquisition and processing in a variety of scenarios. Compression sensing theory can discard the "redundant" information in sparse or compressible signals, reduce the frequency of signals, save the cost of storage and transmission, and collect data for wireless sensor networks. Positioning and other problems provide a new technical solution. This paper focuses on the application of compression sensing technology in virtual MIMO wireless sensor networks, such as adaptive communication and localization, so as to improve the communication performance of wireless sensor networks and improve the accuracy of node location. Firstly, the theory of compression perception is summarized, and several commonly used algorithms of compression perception reconstruction are described in detail. In view of the shortcomings of the traditional compression sensing algorithms, such as high complexity, slow convergence speed and low reconstruction accuracy, based on the basic theory of compression perception, The approximate hyperbolic tangent function with better steepness is used to optimize the approximation function of smooth 0l norm reconstruction algorithm (Smoothed 0ln SL0), and the optimization is carried out according to the mixed optimization algorithm, thus an improved SL0 compression sensing algorithm is proposed. Secondly, the improved compressed sensing reconstruction algorithm is applied to adaptive channel state information feedback of virtual MIMO wireless sensor networks, which reduces the amount of channel state information feedback, reduces the channel overhead, and can reduce the feedback rate at the same time. The maximum total rate of the system and the accuracy of feedback information are increased to improve the network performance and provide a good communication guarantee for wireless sensor network positioning. Finally, on the premise of ensuring the accurate communication of wireless sensor network communication system, the improved compression sensing algorithm is combined with RSS positioning technology, the compressed sensing matrix is constructed by RSS, and the improved algorithm is used to reconstruct the signal. Improve the real-time, accuracy and anti-noise of positioning, achieve high accuracy, fast multi-target positioning in wireless sensor networks.
【学位授予单位】：燕山大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP212.9;TN929.5

【相似文献】