Paper Info
Title
Performance of Square-Range Least Squares and Square-Range Least Absolute Deviations for the Self-Localization of Sensor Nodes Using Convex Relaxations.
Abstract
In the self-localization of sensor nodes using distance measurements, the nodes can be located by minimizing the cost function based on the sum of absolute errors or the sum of squared errors. We investigate the performance of these two cost functions solved by the commonly used SDP relaxation and compare their solution accuracy without and with local refinements. The focus is on the performance under additive and multiplicative TOA noise models, with different minimum number of node connections, and in the presence of measurement outliers.
Year
DOI
Venue
2018
10.1109/ICDSP.2018.8631840
DSL
Keywords
Field
DocType
Cost function,Gaussian noise,Additives,Geometry,Maximum likelihood estimation,Distance measurement,Convex functions
Least squares,Square (algebra),Pattern recognition,Multiplicative function,Computer science,Outlier,Algorithm,Regular polygon,Convex function,Least absolute deviations,Artificial intelligence,Gaussian noise
Conference
ISSN
ISBN
Citations 
1546-1874
978-1-5386-6811-5
0
PageRank 
References 
Authors
0.34
0
2
Name
Order
Citations
PageRank
Shanjie Chen1182.79
K.C. Ho21311148.28