文章基本信息

标题：Development of Multipath Mitigation Methods to Improve GNSS Positioning Accuracy Under Urban Environments
本地全文：下载
作者：Yoshihiro FUKUZAKI ; Yasuhiko ONAKA ; Naohiro TADA 等
期刊名称：Bulletin of the Geographical Survey Institute
印刷版ISSN：0373-7160
出版年度：2019
卷号：67
页码：1-10
出版社：Geographical Survey Institute / Ministry of Land, Infrastructure and Transport
摘要：The accuracy of Global Navigation Satellite System (GNSS) positioning is easily degraded under severe environments such as urban street surroundings due to the blockage of satellite signals by tall buildings. One of the causes of GNSS positioning accuracy degradation is the multipath effect, a phenomenon known to occur in urban areas. To improve the positioning accuracy under such environments, we developed three new methods for multipath mitigation by employing promising techniques from previous studies. The first is the application of an elevation cutoff mask model produced from the distribution data of area obstacles causing the blockage of GNSS signals. The distribution at each site is identified from photos of an overhead sky (e.g. fisheye lens photos), and the elevation cutoff mask model for each site is individually produced and applied to eliminate non-line-of-sight satellites. The second is the application of another elevation cutoff mask model produced from the distribution data of buildings identified from a 3D map. The third is a quality check of the observation data utilizing L1 and L2 Doppler observables. If the observation data contain undesirable signals such as multipath signals, the difference between the L1 and L2 Doppler observables will not agree with the difference in the case of only true signals. By using the discrepancy detected with the Doppler observables, the data can be rejected as invalid observations. These three methods allow observers to select the proper satellites with greater certainty and perform more accurate positioning by identifying the observation data to be rejected. This paper presents the development results of the new methods.