Open AccessA faster way to approximately schedule equally divided jobs with preemptions on a single machine by subsequent job importance growth
Author(s) -
Vadim Romanuke,
Oleksandr Popov
Publication year - 2019
Publication title -
vìsnik harkìvsʹkogo nacìonalʹnogo unìversitetu ìmenì v.n. karazìna. serìâ matematične modelûvannâ, ìnformacìjnì tehnologìï, avtomatizovanì sistemi upravlìnnâ
Language(s) - English
Resource type - Journals
eISSN - 2524-2601
pISSN - 2304-6201
DOI - 10.26565/2304-6201-2019-41-09
Subject(s) - computer science , schedule , computation , scheduling (production processes) , job scheduler , flow shop scheduling , job shop scheduling , mathematical optimization , mathematics , algorithm , operating system , cloud computing
The goal of this work is to study whether the input order of the job release dates results in different time of computations in finding an approximate schedule for equally divided jobs with preemptions on a single machine by subsequent job importance growth,. It has been ascertained that the descending job order has a 1 % relative advantage when scheduling more than 200 jobs. With increasing the number of jobs off 1000, the advantage tends to increase. The advantage can grow up to 22%. A maximally possible gain in computation time is obtained in scheduling longer series of bigger-sized job scheduling problems.