面向信号弱感知能力通信芯片的无线室内定位技术研究

Wireless positioning technologies provide high-resolution and low-cost location services for increasing indoor localization requirements. Targeted at wireless indoor positioning systems (WIPSs) built on communication chips with weak measurement capability of signal strength, this project explores the localization potential of this kind of positioning systems, and focuses on the critical techniques pertaining to the implementation of wireless positioning. Some unreliable factors will be taken into account, e.g., this kind of chips can only sense the values of the received minimum discrete output power (RMDOP) of transmitters obtained by receivers, the values are highly dynamical, and vulnerable to multi-path radio propagation. .In this project, we will mainly study the following three types of key problems, 1) the basic theoretical problems, including the minimum attribute reduction model based optimal deployment problem of reference nodes, and the node localization problem by adopting the idea of probability tree diagram. 2) The node localization problems in line of sight (LOS) environments, including the Markov chain based stable state model of RMDOP value and the area range based node localization problem. 3) The node localization problems based on node arrival matrix and pattern recognition in non line of sight (NLOS) environments. Eventually, the wireless localization algorithms which can resist the impacts of multipath effects and obstacles will be implemented..Carrying out the project will have potential benefits for comprehensive studies of weak measurement capability of signal strength based WIPSs, which has the high importance in the theoretical research aspect and high research value for fostering the development of practical and economical WIPSs.

无线定位技术为日益增长的室内定位需求提供了高精度和低成本的位置服务。本项目针对运行在信号感知能力较弱的低成本通信芯片之上的无线室内定位系统，充分考虑接收端仅能感知发送端最小的离散输出功率(RMDOP)值且该值不稳定、RMDOP值极易受到多径效应影响等因素，探讨其实现定位的潜在能力和研究实现定位涉及的关键技术。重点研究以下问题：1) 基于最小属性约简模型的参考节点优化部署问题，以及基于概率树图思想的节点定位问题等基础性的理论问题。2）视距环境下的节点定位问题，包括基于马尔科夫链的RMDOP值的稳态模型，和基于区域范围的节点定位问题等。3）非视距环境下基于节点到达矩阵和模式识别的节点定位问题，最终实现抗多径效应和障碍物影响的无线室内定位方法。本项目的开展和预期成果能丰富基于信号弱感知能力通信芯片的无线定位的理论与方法，对促进实用、低成本的无线室内定位系统具有较重要的理论意义和研究价值。

无线定位技术为日益增长的室内定位需求提供了高精度和低成本的位置服务。本项目针对运行在信号感知能力较弱的低成本通信芯片之上的无线室内定位系统，充分考虑接收端仅能感知发送端最小的离散输出功率(RMDOP)值且该值不稳定、RMDOP值极易受到多径效应影响等因素，探讨其实现定位的潜在能力和研究实现定位涉及的关键技术。针对基于RMDOP值锚节点部署优化问题，建立了面向参考节点部署的最小属性约简模型，提出了基于最小属性优化问题的粒子群优化算法ANDA，实现了锚节点的优化部署。针对大区域部署，采用区域分块的思想，将锚节点在不同网格点上具有不同的共享度作为其权值，提出了一种能实现锚节点部署总体优化的算法WANDA。针对视距环境下面向信号弱感知能力通信芯片的无线定位问题，提出了一种基于Markov链的RMDOP值稳态值估算算法，利用kNN技术和基于优先级的模式匹配方法，设计了信号指纹的动态构建方法，在此基础上，提出了一种基于指纹信号的定位算法。针对非视距环境下面向信号弱感知能力通信芯片的无线定位问题，利用Markov理论提出了一种基于离线方式的节点到达矩阵构建方法，提出了一种基于时域卷积网络的节点移动行为分类模型和一个基于节点移动行为分类类别的节点定位算法。面向信号弱感知能力通信芯片的无线室内定位原型系统，将nRF24LE1作为选定的、具有代表性的信号弱感知能力通信芯片，我们设计了锚节点和待定位节点的电路原理图，嵌入了锚节点和待定位节点的运行程序和通信协议，实现了硬件设备的原型系统。本项目为信号弱感知能力通信芯片的无线定位提供了一种新颖的解决方案，对促进实用、低成本的无线室内定位系统具有较重要的理论意义和研究价值。基于本项目的研究内容，目前已经在国际、国内学术刊物发表论文篇7篇，其中SCI /EI收录 6 篇，包含 4篇中科院 1 区/CCF A类期刊论文；授权 发明专利6项，软件著作权2项；物联网教材1本。