Research On Single AP Indoor Localization Scheme Based On Wi-Fi Network

Due to the great commercial and social value,indoor localization has attracted extensive attention from academia and industry.With the rapid development of Io T and home automation,more and more Wi-Fi nodes are deployed to indoor scenes.In various indoor localization schemes,Wi-Fi-based localization is the most promising one due to the consideration of positioning accuracy and implementation cost.Indoor localization based on single AP has the advantages of lower implementation cost and higher deployment flexibility,which has a broader application perspective in a relatively small indoor environment.This paper analyzes the characteristics of RSSI,RSS and CSI,discusses the generation and extraction procedure of CSI under the OFDM mechanism,investigates the CSI phase error from the perspective of wireless signal reception and then performs phase calibration of raw CSI data.By comparing,it can be found that CSI has the advantage of better reflecting the channel state than RSS and RSSI.Therefore,we conduct the research of single AP indoor localization based on fine-grained CSI.The main contributions are outlined as follow:First,the AOA estimation of line-of-sight signals are achieved on commodity Wi-Fi device with only 3 antennas.Based on OFDM,Intel 5300 NIC exposes CSI of 30 subcarriers and the TOA of signals introduces measurable phase difference across subcarriers.In view of the problems of indoor coherent signals and limited number of antennas on commodity Wi-Fi devices,the spatial smoothing is conducted on the outputs of antenna array and the spatial spectral estimation techniques and the MUSIC algorithm are adopted to perform twodimensional joint estimation of AOA and TOA for the received signals.The introduction of TOA expands the dimension of the signal steering vector equivalently to the product of the number of antennas and the number of subcarriers,which has broken the limited number of antennas on commodity Wi-Fi devices.According to the distribution characteristics of twodimensional joint estimation of different paths,the cluster analysis is applied to identify the AOA estimation of line-of-sight path.Then,the distance estimation scheme using RSS has problems of low ranging accuracy and poor stability.This paper proposes a ranging scheme based on CIR.To attenuate the interference of indoor multipath effect,multipath suppression is implemented by filtering out the CIR components with large delay,and the CIR data of multiple antennas are combined to avoid the contingency of single antenna reception.We investigated the effects of random phase errors and AGC compensation on the CIR amplitudes.It is concluded that the keys to improving the accuracy of indoor ranging model are CSI phase calibration and AGC compensation elimination.In this paper,a ranging model is established by using CIR which has the ability to distinguish multipath clusters after phase calibration and AGC compensation elimination.After the processing mentioned above,CIR represents stable statistical characteristics which can be used as the energy metric of the receiver at a certain distance.The experimental results in indoor environments show that the ranging model established using the processed CIR has been further improved in accuracy and stability.Finally,we have implemented the single-AP indoor localization system combining AOA estimation and ranging based on Intel 5300 NIC.In the positioning stage,localization system receives the real-time CSI data and fuses the AOA estimation of line-of-sight path obtained by the cluster analysis and the distance estimation obtained from the ranging model to determine the targeted location.After that,the performance of the system is evaluated in many aspects and the experimental results show that the single-AP indoor localization system based on the Wi-Fi network achieves an average positioning accuracy of 1.5m in a typical indoor environment.