Show simple item record

dc.contributor.authorVijayagopal, Charulakshmi
dc.date.accessioned2014-03-04T18:30:10Z
dc.date.available2014-03-04T18:30:10Z
dc.date.issued2010-05
dc.identifier.othervijayagopal_charulakshmi_201005_ms
dc.identifier.urihttp://purl.galileo.usg.edu/uga_etd/vijayagopal_charulakshmi_201005_ms
dc.identifier.urihttp://hdl.handle.net/10724/26527
dc.description.abstractWe consider multiprocessor scheduling of real-time systems of periodic or sporadic tasks. Prior to execution, tasks are partitioned onto the processors using the First-Fit Decreasing algorithm. We demonstrate methods of allocating tasks onto processors even without complete information about the task. We propose two new models to answer two general questions: 1. Given the description of a task set, how many processors should our system have to ensure that the actual task set can be partitioned onto the system? 2. Given the description of a task set and a specific number of processors, how large can our total utilization be? Answers to these are useful to system designers who make design decisions before fully developing tasks. A task set is characterized using maximum utilization, umax and utilization binder, in our first model and a cumulative utilization function Ucum(i) in our second model.
dc.languageeng
dc.publisheruga
dc.rightspublic
dc.subjectPartitioned scheduling
dc.subjectMultiprocessor scheduling algorithm
dc.subjectUtilization bounds
dc.subjectNumber of processors
dc.subjectDissertations
dc.subjectTheses
dc.titleImproving partitioned scheduling bounds
dc.typeThesis
dc.description.degreeMS
dc.description.departmentComputer Science
dc.description.majorComputer Science
dc.description.advisorShelby Funk
dc.description.committeeShelby Funk
dc.description.committeeMaria Hybinette
dc.description.committeeHamid Arabnia


Files in this item

FilesSizeFormatView

There are no files associated with this item.

This item appears in the following Collection(s)

Show simple item record