A Decentralized Approach for Nonlinear Prediction of Time Series Data in Sensor Networks. (10th June 2010)

• Record Type: Journal Article
• Authors: Honeine Honeine, Paul Paul; Richard Richard, Cédric Cédric; Bermudez Bermudez, José Carlos M. José Carlos M.; Chen Chen, Jie Jie; Snoussi Snoussi, Hichem Hichem
• Other Names: Wang Wang Xinbing Xinbing Academic Editor.
• Is Part Of: EURASIP journal on wireless communications and networking. Volume 2010(2010)
• Publication Date: 2010-06-10
• Subjects: Wireless communication systems -- Periodicals; Signal processing -- Periodicals; 621.38405
• DOI: 10.1155/2010/627372
• Languages: English
• ISSNs: 1687-1499

Abstract

Abstract : Wireless sensor networks rely on sensor devices deployed in an environment to support sensing and monitoring, including temperature, humidity, motion, and acoustic. Here, we propose a new approach to model physical phenomena and track their evolution by taking advantage of the recent developments of pattern recognition for nonlinear functional learning. These methods are, however, not suitable for distributed learning in sensor networks as the order of models scales linearly with the number of deployed sensors and measurements. In order to circumvent this drawback, we propose to design reduced order models by using an easy to compute sparsification criterion. We also propose a kernel-based least-mean-square algorithm for updating the model parameters using data collected by each sensor. The relevance of our approach is illustrated by two applications that consist of estimating a temperature distribution and tracking its evolution over time.