Maximum Positioning Error Estimation Method for Detecting User Positions with Unmanned Aerial Vehicle based on Doppler Shifts


IEICE TRANSACTIONS on Communications   Vol.E103-B   No.10   pp.1069-1077
Publication Date: 2020/10/01
Publicized: 2020/05/08
Online ISSN: 1745-1345
DOI: 10.1587/transcom.2019CBP0007
Type of Manuscript: Special Section PAPER (Special Section on New Era of Satellite Communication/Broadcasting/Application Technologies)
unmanned aerial vehicle,  Doppler shift,  positioning accuracy index,  maximum positioning error estimation method,  curving flight model,  

Full Text: FreePDF

In the typical unmanned aircraft system (UAS), several unmanned aerial vehicles (UAVs) traveling at a velocity of 40-100km/h and with altitudes of 150-1,000m will be used to cover a wide service area. Therefore, Doppler shifts occur in the carrier frequencies of the transmitted and received signals due to changes in the line-of-sight velocity between the UAVs and the terrestrial terminal. By observing multiple Doppler shift values for different UAVs or observing a single UAV at different local times, it is possible to detect the user position on the ground. We conducted computer simulations for evaluating user position detection accuracy and Doppler shift distribution in several flight models. Further, a positioning accuracy index (PAI), which can be used as an index for position detection accuracy, was proposed as the absolute value of cosine of the inner product between two gradient vectors formed by Doppler shifts to evaluate the relationship between the location of UAVs and the position of the user. In this study, a maximum positioning error estimation method related to the PAI is proposed to approximate the position detection accuracy. Further, computer simulations assuming a single UAV flying on the curved routes such as sinusoidal routes with different cycles are conducted to clarify the effectiveness of the flight route in the aspects of positioning accuracy and latency by comparing with the conventional straight line fight model using the PAI and the proposed maximum positioning error estimation method.