Cloud-Assisted Cooperative Localization for Vehicle Platoons: A Turbo Approach

Due to the high resolution of angles of arrivals (AoAs) provided by the massive MIMO base station in 5 G wireless systems, it is promising to integrate 5G-based localization technology into autonomous driving to improve the accuracy and robustness of vehicle localization. In this paper, we investiga...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on signal processing 2020, Vol.68, p.605-620
Main Authors: Liu, An, Lian, Lixiang, Lau, Vincent, Liu, Guanying, Zhao, Min-Jian
Format: Article
Language:English
Subjects:
Algorithms
Bayesian analysis
Cloud-assisted cooperative localization
Computer simulation
Localization
Message passing
Statistical inference
Turbo compressive sensing
Vehicle platoon
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Due to the high resolution of angles of arrivals (AoAs) provided by the massive MIMO base station in 5 G wireless systems, it is promising to integrate 5G-based localization technology into autonomous driving to improve the accuracy and robustness of vehicle localization. In this paper, we investigate the problem of 5G cloud-assisted cooperative localization for vehicle platoons. The existing 5G-based localization algorithms focused on single-user localization and are not efficient for the localization of vehicle platoon where the positions of the vehicles are highly correlated. To the best of our knowledge, cloud-assisted cooperative localization tailored to vehicle platoons has not been studied before. To address this challenging problem, we first propose a Gamma-Markov-Group-Sparse (GMGS) model to capture the joint distribution of the vehicle positions in a vehicle platoon. Then we formulate the vehicle platoon cooperative localization as a sparse Bayesian inference (SBI) problem. The existing standard SBI algorithms such as variational Bayesian inference (VBI) and approximate message passing (AMP) cannot be applied to our platoon localization problem due to the complicated GMGS prior and the ill-conditioned measurement matrix. As such, we propose a novel turbo vehicle platoon cooperative localization (Turbo-VPCL) algorithm to fully exploit the correlations of the vehicle positions (as captured by the GMGS prior) under the ill-conditioned measurement matrix. Simulation results verify that the proposed Turbo-VPCL can achieve significant gain over the-state-of-art SBI algorithms.
ISSN:1053-587X
1941-0476
DOI:10.1109/TSP.2020.2964198