Title

Load balancing as cooperative problem solving in distributed artificial intelligence

Date of Publication

1995

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

Jefferson Tan

Defense Panel Chair

Kelsey Hartigan Go

Defense Panel Member

Dr. Lloyd Espiritu
Dr. Harry Joson

Abstract/Summary

In any distributed system, there are processors wherein the computational capacities are too small to take enormous amounts of time such that the response time and output of these computations may not be reasonable. One solution to this situation is through balancing the system workload. Load balancing allows remote execution of users' tasks even in the absence of idle processors and at the same time, strives to equalize the system workload among all the processors in a distributed system. Achieving this requires cooperation among processors in the system. In distributed artificial intelligence (DAI), a model known as the cooperative problem solving (CPS) process has been proposed [WOOL94b] for modelling interactions among a group of logically decentralized agents that choose to work together to achieve a common goal. This research presents a load balancing algorithm using the 4 stages that comprise the CPS model. It was shown that the load balancing process is an instance of CPS. A formal model of this algorithm was presented using quantified modal logic. The algorithm performs load balancing globally and takes into consideration the processors' load and resources.

Abstract Format

html

Language

English

Format

Print

Accession Number

TG02419

Shelf Location

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

Physical Description

219 leaves; 28 cm.

Keywords

Artificial intelligence; Problem solving; Electronic data processing -- Distributed processing; Algorithm; Digital computer simulation

This document is currently not available here.

Share

COinS