【作者】 艾中进；

【导师】 刘晔；

【作者基本信息】 南昌大学 ， 信号与信息处理， 2010， 硕士

【摘要】 射频识别(Radio Frequency Identification, RFID)是一种利用射频信号自动识别目标对象并获取相关信息的技术,识别工作可自动完成无须人工干预,相关设备可工作于各种恶劣环境下。基于RFID的定位系统在国内外都有着广泛的应用,目前RFID定位系统常用的定位方法有场强定位和到达时间差定位,对到达信号的时延估计是进行到达时间差定位的前提。本文作者对目前几种常用的时延估计算法做了研究,建立了仿真模型分析其性能,并寻找提高算法性能的突破口。本文的主要研究工作如下：1、介绍了RFID定位系统的组成及其工作原理,国内外的发展现状和主要应用；同时分析了几种主要的无线定位技术的工作原理,以及影响定位精度的主要因素。2、分析了互相关算法的原理,对互相关算法、Roth脉冲响应处理器法、PHAT互功率谱处理器法和SCOT平滑相干变换器算法进行分析,并对四种时延估计方法做了仿真,仿真结果表明,加权的互相关算法时延估计的正确率高于互相关算法。3、对互相关算法采用了二次多项式最小二乘拟合法来改进,算法改进的思路是将时延估计分两步：先采用广义互相关算法对时延进行初步的估计,再在初步估计的基础上采用二次多项式最小二乘拟合法,对时延进行二次估计,以进一步提高时延估计的精度。更多还原

【Abstract】 Radio Frequency Identification (RFID) is a technology, which using radio waves to automatically collect relevant information, and the working of collection can be done without human’s intervention, and the relevant equipment can be worked in a variety of harsh environments. The RFID location systems are wildly used in the world. RFID location systems usually use two methods to locate the object; they are Received Signal Strength Indicator (RSSI) and Time Difference of Arrival (TDOA). Estimating the delay time of arrival signal is the premise of using time difference of arrival to locate.To improve the precision of time delay estimation, this dissertation studies time delay estimation algorithm and builds time delay simulating model. The main contents are as follows:This dissertation describes the working principle of RFID location system, and the main application at home and abroad. This dissertation introduces several major wireless location technologies, and analyzes the main factors of affecting the locating accuracy.This dissertation analyzes the principles of basic correlation algorithms, and compares the basic correlation algorithms, Roth Impulse Response Processor (Roth), Phase Transform (PHAT), and Smooth Coherence Transform (SCOT). All of the four algorithms are simulated in Matlab.Using the method of least squares quadratic polynomial fitting improves the accuracy of correlation algorithm. The idea of improving time delay estimation algorithm can be divided into two steps, firstly to get the coarse data using the cross correlation algorithm, then to obtain better results with least squares quadratic polynomial fitting method.更多还原