Non-preemptive throughput maximization for speed-scaling with power-down

Abstract : We consider the problem of scheduling a set of n jobs on a single processor. Each job is characterized by its release date rj, its deadline dj and its processing volume pj. The processor can vary its speed and can switch into a sleep state in order to reduce its energy consumption. No energy is consumed in this state, but a fixed amount of energy, equal to L, is required for a transition from the sleep state to the active state. Here, we study the throughput maximization version of the problem where we are given a budget of energy E and our goal is to determine a feasible schedule maximizing the number of jobs that are executed between their respective release dates and deadlines without preemption. We first consider the case in which jobs have agreeable deadlines, i.e. for every pair of jobs i and j, one has ri ≤ rj if and only if di ≤ dj. Then we consider the case where the jobs have arbitrary release dates and deadlines, but the same processing volume. We propose polynomial-time algorithms for both cases.
Document type :
Conference papers
Liste complète des métadonnées

https://hal.archives-ouvertes.fr/hal-01221389
Contributor : Frédéric Davesne <>
Submitted on : Tuesday, October 27, 2015 - 11:07:02 PM
Last modification on : Thursday, March 21, 2019 - 12:59:46 PM

Identifiers

Citation

Eric Angel, Evripidis Bampis, Vincent Chau, Kim Thang Nguyen. Non-preemptive throughput maximization for speed-scaling with power-down. 21st International Conference on Parallel and Distributed Computing (Euro-Par 2015), Aug 2015, Vienna, Austria. pp.171--182, ⟨10.1007/978-3-662-48096-0_14⟩. ⟨hal-01221389⟩

Share

Metrics

Record views

246