Task-driven automated component deployment for ambient intelligence environments

  • Authors:
  • Peter Rigole;Tim Clerckx;Yolande Berbers;Karin Coninx

  • Affiliations:
  • Department of Computer Science, K.U. Leuven, Celestijnenlaan 200A, BE-3001 Leuven, Belgium;Hasselt University, Expertise Centre for Digital Media, and Transnational Universiteit Limburg, Wetenschapspark 2, BE-3590 Diepenbeek, Belgium;Department of Computer Science, K.U. Leuven, Celestijnenlaan 200A, BE-3001 Leuven, Belgium;Hasselt University, Expertise Centre for Digital Media, and Transnational Universiteit Limburg, Wetenschapspark 2, BE-3590 Diepenbeek, Belgium

  • Venue:
  • Pervasive and Mobile Computing
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

This article presents a strategy for deploying component-based applications gradually in order to match the functionality of pervasive computing applications onto the current needs of the user. We establish this deployment strategy by linking component composition models with task models at design-time, from which a run-time deployment plan is deduced. Enhanced with a Markov model, this deployment plan is able to drive a component life cycle manager to anticipate future deployments. The result is a seamless integration of pervasive computing applications with the user's tasks, guaranteeing the availability of the required functionality without wasting computing resources on components that are not currently needed.