A server repeatedly transmits data items (pages) possibly with different speeds on a set of channels. The objective is to minimize energy consumption of the schedule. We adopt the common model that sending at speed s for t time units consumes t · sα energy for a constant α ≥ 2. An individual window length is associated with each page. This length is a strict upper bound on the time between two consecutive broadcasts for that page. We present an easy to implement algorithm for the single channel case that obtains an approximation ratio of 3 · 4α. For the multi-channel case with identical channels an extension of this algorithm computes an 8α-approximation. Both algorithms run in low-order polynomial time. As our main tool for the analysis, we show that it suffices to consider periodic schedules as their energy density (total energy consumption per time unit) differs from the one of general schedules at most by (1 + ε) for an arbitrary constant ε > 0. © Springer-Verlag Berlin Heidelberg 2008.
CITATION STYLE
Gunia, C. (2008). Energy-efficient windows scheduling. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 4910 LNCS, pp. 304–315). Springer Verlag. https://doi.org/10.1007/978-3-540-77566-9_26
Mendeley helps you to discover research relevant for your work.