We have the pleasure to receive the visit of Stéphan Plassart (EPFL) on May 9th (14:30, A115 at Inria, Paris) when he gives a talk entitled Online energy optimization using dynamic speed scaling in real-time systems.
You may join us by visio French only
Abstract:
The energy consumption is a crucial issue for real-time systems, that’s why optimizing it online, while the processor is running, has become essential. This optimization can be done by adapting the processor speed during the job execution. In this talk, we will use dynamic programming to compute the optimal speed scaling policy that minimizes the energy consumption of a single processor executing a finite or infinite set of jobs with real-time constraints. Several situations will be studied with different knowledge on past, active and future job characteristics. Firstly, we consider that all job characteristics are known (the offline case), and we propose a linear time algorithm to determine the speed schedule to execute n jobs on a single processor. Secondly, we propose a Markov decision process (MDP) approach that solves the case where past and active job characteristics are entirely known, and for future jobs only the probability distribution of the job characteristics (arrival times, execution times and deadlines) is known. To finish, we also consider the case where we have no statistical knowledge on jobs, so we have to use learning methods to determine the optimal processor speeds online.
Bio:
Stéphan Plassart is a postdoctoral researcher at École polytechnique fédérale de Lausanne (EPFL) in Switzerland in the LCA2 team of Professor Jean-Yves Le Boudec. He completed his PhD at Grenoble Alpes University, in the POLARIS and SPADES teams at LIG and at Inria Grenoble under the supervision of Bruno Gaujal and Alain Girault. His current research focuses on network calculus for Time-Sensitive Network (TSN). His research interests are also in optimization, Markov decision process, learning, energy, and real-time systems.