【作者】 邵成刚；

【导师】 景晓军；

【作者基本信息】 北京邮电大学 ， 通信与信息系统， 2013， 硕士

【摘要】 现阶段,随着移动通信以及室内无线技术的快速发展,室内定位技术也成为了人们探索和研究的热门课题之一。基于室内无线技术的室内定位系统具有很好的商业价值和工业前景,但是如红外线、超声波、GPS等典型的室内定位技术,现在已经无法满足高精度,低成本的要求,而射频识别(RFID)技术有成本低,非视距,非接触,定位性能好等特点,使它能在室内定位领域中成为了佼佼者。虽然现阶段已存在不少经典的基于RFID的室内定位系统或者算法,但是由于室内环境的复杂性,射频信号之间的多径干扰等使得对高抗干扰性,高定位精度的室内定位算法提出了新的要求。本课题首先对用于室内定位的关键技术进行概述,然后对经典室内定位算法进行了分析,如Fang算法、LANDMARC算法、VIRE算法,介绍了它们各自的优缺点。在上述的基础上,从Reference Tag的布局,Virtual Reference Tag插值法和“最邻近”Tag的权值定义法三个方面进行了探索和改进。第一,从概率和几何上,分析了矩形Tag布局的缺点,提出了三角形Tag布局,在减少Tag的数量的同时定位算法的精度提高了3.4%；第二,结合室内环境的多变性和二维牛顿插值法,提出了新的Virtual Reference Tag插值函数。相比于线性插值法和拉格朗日插值法,平均定位精度提高27.4%和9.8%,并且增加了定位系统的稳定性；最后,根据无线信道的传播模型,提出了基于“距离—损耗”公式的权值定义法,分析仿真结果定位精度有2.1%的提高。综上所述,通过仿真结果证明,本文在基于RFID室内定位算法三个方面的改进使系统的定位精度和稳定性比LANDMARC算法和VIRE算法有了显著的提高。更多还原

【Abstract】 At this stage, with the rapid development of mobile communication and indoor wireless technology, indoor positioning technology has also become one of the hot topics for people to explore and research. Indoor positioning system based on indoor wireless technology has good commercial value and industrial prospects, but typical indoor positioning technologies such as infrared, ultrasonic, GPS, now unable to meet the requirements of high-precision, and radio frequency identification (RFID) have advantage of low-cost, non-line-of-sight, non-contact positioning performance, so that it can become a leader in the field of indoor positioning. Although in recent years there are many classic RTTD-based indoor positioning system or algorithm, but due to the complexity of the indoor environment, multipath interference between the RF signals makes indoor positioning algorithm for high noise immunity and high positioning accuracy the new requirements.At first, we give an overview of the key technologies for indoor positioning, and do analysis of the classic indoor positioning algorithm, Fang algorithm, LANDMARC algorithm, VIRE algorithm, and introduce their respective advantages and disadvantages. On the basis of the above, we explore and improved from three aspects on the layout of the Reference Tag, Virtual Reference Tag interpolation and the nearest Tag weights defined law. Firstly, we analyse the shortcomings of rectangular Tag layout through probability and geometry, put forward a triangle Tag layout, and the accuracy of the positioning algorithms improve3.4%with reducing the number of Tag. Secondly, combined with the variability of the indoor environment and two-dimensional Newton interpolation method we proposed a new the Virtual Reference Tag interpolation function. The average positioning accuracy increased by27.4%and9.8%compared to the linear interpolation and Lagrange interpolation method, and it also increases the stability of the positioning system. Finally, according to the radio channel propagation model, according to the analysis of simulation, the formula weights defined method based on the distance-lossa curve results2.1%increase on positioning accuracy.In summary, the system’s positioning accuracy and stability has been significantly improved through the improvement based on the three aspects of the RFID indoor positioning algorithm in this paper than LANDMARC algorithm and VIRE algorithm through simulation results proved.更多还原