Home // COGNITIVE 2016, The Eighth International Conference on Advanced Cognitive Technologies and Applications // View article

Applying Pairing Support Vector Regression Algorithm to GPS GDOP Approximation

Authors:
Pei-Yi Hao
Chao-Yi Wu

Keywords: geometric dilution of precision (GDOP); Global Positioning System (GPS); kernel-based method; support vector machine; support vector regression.

Abstract:
Global Positioning System (GPS) has extensively been employed in various applications. Geometric Dilution of Precision (GDOP) is an indicator showing how well the constellation of GPS satellites is geometrically organized. GPS positioning with a smaller GDOP value usually yields better accuracy. However, the calculation of GDOP is a time- and power-consuming task that requires to solving measurement equations with complicated matrix transformation and inversion. When selecting the one with the lowest GDOP for positioning from many GPS constellations, methods that can fast and accurately calculate GPS GDOP are imperative. Previous works have shown that numerical regression on GPS GDOP can yield satisfactory results and eliminate many calculation steps. This paper employs a new pairing support vector regression algorithm (pair-SVR) to the approximation of GPS GDOP. The pair-SVR determines indirectly the regression function through a pair of nonparallel insensitive upper- and lower-bound functions, each of which is solved by support vector machine (SVM)- type quadratic programming problems (QPP) with smaller-sized. This strategy makes the pair-SVR not only have the faster learning speed than the classical SVR, but also be suitable for many cases, especially when the noise is heteroscedastic. Besides, pair-SVR improves the sparsity than that of TSVR by employing the concept of insensitive zone. This makes the prediction time complexity of pair-SVR is obviously smaller than TSVR. The experimental results show that pair-v-SVR gains better performance for the approximation of GPS GDOP than previous support vector regression machine.

Pages: 102 to 107

Copyright: Copyright (c) IARIA, 2016

Publication date: March 20, 2016

Published in: conference

ISSN: 2308-4197

ISBN: 978-1-61208-462-6

Location: Rome, Italy

Dates: from March 20, 2016 to March 24, 2016