Online packing with gradually improving capacity estimations and applications to network lifetime maximization

  • Authors:
  • Marcel Ochel;Klaus Radke;Berthold Vöcking

  • Affiliations:
  • Department of Computer Science, RWTH Aachen University, Germany;Department of Computer Science, RWTH Aachen University, Germany;Department of Computer Science, RWTH Aachen University, Germany

  • Venue:
  • ICALP'12 Proceedings of the 39th international colloquium conference on Automata, Languages, and Programming - Volume Part II
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

We introduce a general model for online packing problems with applications to lifetime optimization of wireless sensor networks. Classical approaches for lifetime maximization make the crucial assumption that battery capacities of sensor nodes are known a priori. For real-world batteries, however, the capacities are only vaguely known. To capture this aspect, we introduce an adversarial online model where estimates become more and more accurate over time, that is, when using the corresponding resources. Our model is based on general linear packing programs and we assume the remaining capacities to be always specified by lower and upper bounds that may deviate from each other by a fixed factor α. We analyze the algorithmic consequences of our model and provide a general -competitive framework. Furthermore, we show a complementary upper bound of.