| High-precision positioning has always been one of the technologies that people are willing to explore in daily life.In the past of decades,outdoor positioning technology based on satellite and mobile networks has gradually become a unified standard,while indoor positioning is in urgent need of further development and has become a research hotspot.The scheme of indoor positioning using RSSI of Wi-Fi,Bluetooth and other nodes also has a relatively complete theoretical framework.However,for Wi-Fi signals,its Channel State Information can characterize the channel characteristics in a more subtle way,and the scheme using CSI for positioning has not yet been implemented.This paper built a set of experimental extraction platform for CSI data,and designed a set of experimental methods to achieve high-precision distance estimation.The experiments were carried out in ideal environment and indoor environment.Through the analysis of its CSI information,a phase-based time-of-flight estimation method was explored.The platform was implemented based on Intel 5300 network card.Two-stage fitting method has been proposed,which can effectively solve the problem that it is difficult to achieve correct fitting due to the large number of parameters,thereby greatly improving the accuracy of nonlinear fitting of the CSI model.Experiments show that in the ideal experimental scenario,the improved method is used to process the channel state information to estimate the flight time to estimate the distance.In the ideal state,the degree of dispersion is extremely small and reach centimeter level In the indoor environment,the distance estimation error can reach the decimeter level,the standard deviation at a distance of 8m can is 0.64 m.In the environment affected by non-line-of-sight and multipath,the standard deviation at a distance of 8m can still reach 1.04 m.With the commercialization of 5G technology,the characteristics of 5G high-density networking and the introduction of D2 D make high-precision indoor and outdoor fusion positioning easier.This thesis built a simulation experiment platform,restores the actual scene of the teaching and research building and the layout of Wi-Fi nodes,and realizes the software simulation of the 5G-assisted indoor and outdoor fusion positioning scheme.The simulation platform includes Wi-Fi as the main data source,deploys 5G base stations,introduces PDR and D2 D as fusion data sources to improve the positioning accuracy in the horizontal direction,and introduces barometers to achieve vertical positioning.Based on the above platform,this thesis proposed a set of probability-based particle screening and fusion algorithms,and simulated the extended Kalman filter algorithm commonly used in fusion positioning.The simulation verifies the proposed fusion positioning scheme under different paths.With RMSE as the judgment criteria,the RMSE reaches 1.1441 m in indoor scenarios and 2.9659 m in outdoor scenarios. |
PDF Full Text Request