Projection-based linear constrained estimation and fusion over long-haul links

We study estimation and fusion with linear dynamics in long-haul sensor networks, wherein a number of sensors are remotely deployed over a large geographical area for performing tasks such as target tracking, and a remote fusion center serves to combine the information provided by these sensors in o...

Full description

Saved in:
Bibliographic Details
Published in2016 IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems (MFI) pp. 436 - 441
Main Authors Qiang Liu, Rao, Nageswara S. V.
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.09.2016
Subjects
Covariance matrices
Dynamics
error covariance matrices
linear constraints
Long-haul sensor networks
projection method
root-mean-square-error (RMSE) performance
State estimation
state fusion
Target tracking
Time measurement
Online AccessGet full text

Cover

More Information
Summary:We study estimation and fusion with linear dynamics in long-haul sensor networks, wherein a number of sensors are remotely deployed over a large geographical area for performing tasks such as target tracking, and a remote fusion center serves to combine the information provided by these sensors in order to improve the overall tracking accuracy. In reality, the motion of a dynamic target might be subject to certain constraints, for instance, those defined by a road network. We explore the accuracy performance of projection-based constrained estimation and fusion methods that is affected by information loss over the long-haul links. We use an example to compare the tracking errors under various implementations of centralized and distributed projection-based estimation and fusion methods and demonstrate the effectiveness of using projection-based methods in these settings.
DOI:10.1109/MFI.2016.7849527