Home // SENSORCOMM 2012 , The Sixth International Conference on Sensor Technologies and Applications // View article

Classification of Faults in Sensor Readings with Statistical Pattern Recognition

Authors:
Valentina Baljak
Kenji Tei
Shinichi Honiden

Keywords: Fault Tolerance; Wireless Sensor Networks; Pat- tern Recognition; Faults Classification

Abstract:
In wireless sensor networks, frequent faults are caused by general characteristics and the direct exposure to the environment. Accumulation of these faults can lead to the progressive decrease of reliability and accuracy of sensor readings. We focus on detection and classification of faults within sensory data independently of the underlying cause. We propose a complete and consistent fault classification based on two aspects. The first aspect is continuity and frequency of the occurrence, and the second is the existence of observable and learnable patterns. Given that modeling of faults prior to the detection is a fundamental process, we address it with statistical analysis and theoretical approach. We rely on centralized and straightforward detection methods using neighborhood vote. For the full classification phase, we propose the use of statistical pattern recognition with a priori modeling of faults. Current results show that this method works comparatively well when applied to collected data in data centric dense wireless sensor network.

Pages: 270 to 276

Copyright: Copyright (c) IARIA, 2012

Publication date: August 19, 2012

Published in: conference

ISSN: 2308-4405

ISBN: 978-1-61208-207-3

Location: Rome, Italy

Dates: from August 19, 2012 to August 24, 2012