Design And Implementation Of A Single Node Indoor Localization System Based On Configurable Antennas

With the development of wireless communication technologies and the Internet,the application scenarios of positioning technologies in people's lives are becoming more and more abundant.Therefore,people's demand for location-based services is also increasing greatly.Besides,due to the widespread deployment of commercial WIFI devices,the construction of WIFI-based localization system is very promising.This dissertation focuses on the localization algorithms utilizing Channel State Information(CSI)in the single-node indoor localization system,and studies the impact of different external antenna deployments on the performance of the system.Finally,an indoor positioning system with one single WIFI node is built.and tested in various indoor scenario to verify the performance of the system.First of all,this article introduces the development and application prospects of the current indoor localization technology and describes the characteristics and deficiencies of several existing indoor localization technologies.The WIFI based indoor localization technology is introduced in detail.The theory and development of fingerprint location algorithm based on received signal strength indicator(RSSI)are analyzed,and then introduce the research status of CSI-based WIFI localization.After that,this paper analyzes the theoretical basis and method of WIFI localization and then the CSI acquiring method based on CSITool toolkit as well as the theory of localization using CSI are introduced.Then,the error introduced in the signal processing of the WIFI signal in the receiver is studied,and a nonlinear phase error correction method based on overlapping subcarriers as well as a linear phase error correction method based on a linear model are proposed.Then,based on the corrected CSI,a band splicing method for overlapping subcarriers is proposed herein,which is respectively combined with localization algorithms using power delay profile(PDP)and multiple signal classification algorithm(MUSIC).The combination of localization algorithms can improve the positioning accuracy of the original algorithm.Utilizing the stability of the direct signal transmission path,a cluster-based direct path identify algorithm is proposed,which can further improve the localization performance.This paper also analyzes and compares the effects of the localization performance of different antenna deployment methods,i.e.linear antenna arrays and plane antenna array,both theoretically and practically.Finally,this paper builds a single-node indoor positioning system based on Linux platform.Under the configuration of linear array antenna and area array antenna,the performance of different indoor scenes is tested respectively.According to the analysis of localization results,the localization accuracy of the single-node localization system established in the ideal environment is 1.3m when the confidence coefficient is 50%.It shows that the effectiveness of the proposed localization algorithms and the positioning accuracy under plane antenna array configuration is better than linear antenna array.