Paper Info
Title
GNSS Vector Tracking Method Using Graph Optimization
Abstract
Commonly, there are two different types of signal tracking methods in Global Navigation Satellite System (GNSS) receivers: Scalar Tracking (ST) and Vector Tracking (VT). In ST, each tracking channel works independently and generates the measurements for Position, Velocity and Time (PVT) estimation. On the contrary, VT accomplishes the signal tracking and PVT solutions estimation together using a Kalman filter (KF). In this manner, mutual aiding between these channels are obtained, which contributes to the superior performance of VT than ST. KF is the method in the manner of weighted average of predicted state and the measurements derived state. Two drawbacks restrict the KF performance, (1) in KF, correlation between current state and all the past states is ignored, only adjacent state is included in the current state estimation through state transformation matrix; (2) linear KF might degrade the estimation due to the nonlinearity of the VT measurement model. In this brief, a graph optimization (GO) method is employed to substitute the KF to estimate the navigation solutions in VT. In the GO-VT, the measurements and the state transformation from past epochs are all regarded as constraints to optimize the states estimation. With the iterations during the optimization, the GO-VT is expected to alleviate the influence of the measurement model nonlinearity on the state estimation. A field test was carried out for assessing the performance of the GO-VT, its superior position accuracy compared with that from the KF supported that the GO could enhance the VT.
Year
DOI
Venue
2021
10.1109/TCSII.2020.3025455
IEEE Transactions on Circuits and Systems II: Express Briefs
Keywords
DocType
Volume
GNSS,graph optimization,Kalman filter
Journal
68
Issue
ISSN
Citations 
4
1549-7747
1
PageRank 
References 
Authors
0.36
0
5
Name
Order
Citations
PageRank
Changhui Jiang1146.75
Shuai Chen2339.17
Yuwei Chen312923.23
Di Liu441.13
Yuming Bo55411.62