salle A. Turing CE4
30 January 2014 - 15h00
Multi-objective scheduling for multi-core systems
by Abhinav Srivastav from Verimag
Abstract: The aim of this thesis is to study the scheduling problems in context of multi-objectives. Most of the problem pertaining to scheduling on multi-core machines are NP hard. Therefore, we are interested in finding the good approximations for such problems. One such problem that we consider here is the construction of pareto optimal front for jobshop problem with stretch for each job as quality of service. In this presentation, I will discuss about the problems pertaining to construction of such pareto approximation set and methodology adapted by us to design a fast search algorithm for above problem. Apart from this problem, we has also focus our work on reducing average stretch for streaming jobs(non-preemptive schedule). I will discuss an algorithm which provides good bound on competitive ratio for this problem.