| With the rapid development of the mobile Internet era,users' work and life are more completed indoors.Indoor positioning has become an increasingly important technology and service.Wi-Fi-based indoor positioning of mobile terminals is of great social value and scientific significance.Considering that the use of fingerprint positioning to achieve indoor positioning requires large human resources and low positioning accuracy,this paper mainly studies the positioning method by acquiring signal delay to achieve ranging.The main research contents and contributions of this article are as follows:Through the search and study of the literature,we understand and introduce the research significance of indoor positioning based on Wi-Fi signal as a signal source and the current research status at home and abroad,and introduce the overall structure of the article.The Wi-Fi signal structure and characteristics in IEEE 802.11 are described and introduced in detail,as well as the characteristics and composition of the long and short training sequences used in time delay estimation.Finally,we briefly introduced several positioning methods based on TOA,including commonly used circumferential positioning models,hyperbolic positioning models,and improved positioning models such as Caffery,Chan,and AML.After the Wi-Fi signal propagates in the wireless channel,the signal needs to be preprocessed,including the use of long and short training sequences for signal detection,frequency offset correction,signal alignment,and channel estimation.The signal alignment section introduces continuous broadcast based Two solutions for continuous signal transmission.The signal estimation algorithm introduces three estimation methods: LS channel estimation method,LMMSE channel estimation method,DFT,etc.,and analyzes the accuracy and advantages and disadvantages of the three algorithms through simulation experiments.Aiming at the problem that multipath delay cannot be resolved in a multipath channel environment.The matrix transformation delay estimation algorithm separately describes four super-resolution delay estimation algorithms such as ESPRIT,TLSESPRIT,MUSIC,ROOT-MUSIC.Through simulation experiment analysis,it is known that the super-resolution delay estimation algorithm based on MUSIC has the best performance.And through the smoothing algorithm to solve the problem of the MUSIC delay estimation algorithm in the case of a single snapshot,the super-resolution delay estimation algorithm based on compressed sensing,this article introduces the MP estimation algorithm,OMP estimation algorithm,ROMP estimation algorithm and detailed Describes the flow of the algorithm and explains the basis for applying the algorithm to the delay estimation problem.For the super-resolution delay estimation algorithm,a delay grid dictionary needs to be built,which leads to the problem that the actual delay does not match the estimated delay.Therefore,this paper proposes a twostep estimation super-resolution delay algorithm.The MUSIC algorithm or OMP algorithm The signal is roughly estimated,and then the spherical interpolation algorithm and the differential channel power method are used for further accurate estimation,thereby effectively solving the problem that the actual delay does not match the estimated delay.The grid limitation is improved,and the effectiveness and accuracy of the proposed algorithm are analyzed through simulation experiments.This paper rewrites the algorithm proposed above through NI-USRP(National Instrument-Universal Software Radio Peripheral,NI-USRP)platform using LabVIEW language,successfully implements a real-time test system,and verifies the algorithm through field experiments Ranging experiments of 5m and 12 m were carried out,and the results were obviously better than traditional estimation algorithms. |
