A Methodology for Modelling Mobile Agent-Based Systems (Mobile agent Mobility Methodology - MaMM)

Mobile agents are a particular type of agents that have all the characteristics of
an agent and also demonstrate the ability to move or migrate from one node to
another in a network environment. Mobile agents have received considerable
attention from industry and the research community in recent times due to the
fact that their special characteristic of migration help address issues such as
network overload, network latency and protocol encapsulation. Due to the current
focus in exploiting agent technology mainly in a research environment, there has
been an influx of software engineering methodologies for developing multi-agent
systems. However, little attention has been given to modelling mobile agents. For
mobile agent-based systems to become more widely accepted there is a critical
need for a methodology to be developed to address various issues related to
modelling mobility of agent . This research study provides an overview of the
current approaches, methodologies and modelling languages that can be used
for developing multi-agent systems. The overview indicates extensive research
on methodologies for modelling multi-agent systems and little on mobility in
mobile agent-based systems. An original contribution in this research known as
Mobile agent-based Mobility Methodology (MaMM) is the methodology for
modelling mobility in mobile agent-based systems using underlying principles of
Genetic Algorithms (GA) with emphasis on fitness functions and genetic
representation. Delphi study and case studies were employed in carrying out this
research.