PC3: Principal Component-based Context Compression

  • Authors:
  • Christos Anagnostopoulos;Stathes Hadjiefthymiades;Panagiotis Georgas

  • Affiliations:
  • Department of Informatics, Ionian University, Corfu, 49100, Greece;Department of Informatics and Telecommunications, University of Athens, Athens, 15784, Greece;Department of Informatics and Telecommunications, University of Athens, Athens, 15784, Greece

  • Venue:
  • Journal of Parallel and Distributed Computing
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

We focus on energy efficiency, which guarantees the operation of a Wireless Sensor Network for long. We propose a context compression model that works in an orthogonal fashion. We first reduce the dimensions of multivariate contextual information. This is achieved through the Principal Component Analysis (PCA), which determines the statistical dependencies between the different contextual components. We then suppress the transmission of the determined principal components through an extrapolation scheme that exploits the properties of each individual component. Our findings are quite promising for the broader domain of WSN-based application engineering and context awareness.