Scalability and Latency Analysis of the Centralized 3D Mapping Aided GNSS-Based Collaborative Positioning

Guohao Zhang, Hoi-Fung Ng, Li-Ta Hsu, Han Gao, and Dingzhong Yao

Abstract:	The recent development of the collaborative positioning (CP) algorithm enables a new possibility to improve the positioning performance, by making use of the measurements from neighboring road agents. Among different CP algorithms, the 3D mapping aided (3DMA) GNSS based CP has shown great potential to improve the GNSS positioning performance for the road agents located in the urban canyon. However, due to the difficulty of conducting real experiments, the 3DMA GNSS CP is usually only validated by the collaboration between few agents. The influence of the collaboration network size on the CP performance is worth to be investigated before its practical applications. Besides, the communication latency during CP may also degrade the positioning performance in real applications. In this study, based on a realistic GNSS measurement simulator, the impacts of the network size and the communication latency on the performance of the centralized 3DMA GNSS CP algorithm are investigated. The scalability analysis result shows that, simply enlarging the network size may not always improve the CP performance, or even worse, degrades the performance by involving additional error sources. On the other hand, the CP performance can be sensitively degraded by communication latency, which needs to be mitigated in practical applications.
Published in:	Proceedings of the 2021 International Technical Meeting of The Institute of Navigation January 25 - 28, 2021
Pages:	307 - 316
Cite this article:	Zhang, Guohao, Ng, Hoi-Fung, Hsu, Li-Ta, Gao, Han, Yao, Dingzhong, "Scalability and Latency Analysis of the Centralized 3D Mapping Aided GNSS-Based Collaborative Positioning," Proceedings of the 2021 International Technical Meeting of The Institute of Navigation, January 2021, pp. 307-316. https://doi.org/10.33012/2021.17835
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In