Date of Publication

6-25-1992

Document Type

Master's Thesis

Degree Name

Master of Science in Computer Science

Subject Categories

Computer Sciences

College

College of Computer Studies

Department/Unit

Computer Science

Thesis Adviser

En-Hsin Huang

Defense Panel Chair

Mitch Andaya

Defense Panel Member

Kelsey Hartigan Go
Peter Fernandez

Abstract/Summary

In a Distributed Computing System (DCS) jobs can arrive randomly at each node, which can change the status of the node constantly. Therefore, jobs in DCS should be scheduled dynamically to meet the constraints of the system and to improve the system performance. For job scheduling, accurate global information is impossible. However, an estimation can be made to schedule job to achieve near-optimal solution of the problem of job scheduling. For reliability, a scheduler should be placed on each node in the system.

This study is focuses on dynamic job scheduling in DCS using network of stochastic learning automata (SLA). SLA is used as a decision maker in job scheduling. First, an abstract model of DCS is presented, then the algorithm is formulated for dynamic job scheduling. A mathematical proof of correctness is conducted for the validation of the algorithm.

Abstract Format

html

Language

English

Format

Print

Accession Number

TG02004

Shelf Location

Archives, The Learning Commons, 12F Henry Sy Sr. Hall

Physical Description

146 leaves, illustrations, 28 cm.

Keywords

Stochastic programming; Scheduling (Management); Electronic data processing--Distributed processing; Distributed computer systems in electronic data processing

Upload Full Text

wf_yes

Download

Share

COinS