An agent based approach towards autonomic services
Services play an increasingly important role in software applications today. There are increasing demands to build/compose software as a collection of services. A service in this context may be defined as a behaviour that is provided by a component for use by any other component based on a networ...
Saved in:
Main Author: | |
---|---|
Other Authors: | |
Format: | Theses and Dissertations |
Language: | English |
Published: |
2009
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/14959 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
Summary: | Services play an increasingly important role in software applications today.
There are increasing demands to build/compose software as a collection of services. A
service in this context may be defined as a behaviour that is provided by a component
for use by any other component based on a network-addressable interface contract.
Such service oriented software raises some challenges when designed using current
object-oriented methodologies in situations where there is a need to trade-off
competing goals, where there are complex business workflows, or when service
execution is highly dynamic. Multi-Agent Systems (MAS) fits well into complex
services, but current Agent-Oriented Software Engineering (AOSE) methodologies do
not address the aforementioned challenges of services. AOSE for services that operate
in open, dynamic and complex environments such as Grids is a major problem that
needs to be addressed.
This thesis describes an AOSE methodology that is particularly suited for
building service oriented software. It has the advantages of being accessible and
pragmatic. The thesis presents a novel Goal-Oriented and FCM for Agent-mediated
Autonomic Services (GOFASS) methodology used to specify agents for the agentmediated
autonomic services architecture. The GOFASS methodology provides
guidance from early requirements gathering through to agent design and
implementation. |
---|