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

摘要

Mobile agents are a particular type of agents that have all the characteristics ofudan agent and also demonstrate the ability to move or migrate from one node toudanother in a network environment. Mobile agents have received considerableudattention from industry and the research community in recent times due to theudfact that their special characteristic of migration help address issues such asudnetwork overload, network latency and protocol encapsulation. Due to the currentudfocus in exploiting agent technology mainly in a research environment, there hasudbeen an influx of software engineering methodologies for developing multi-agentudsystems. However, little attention has been given to modelling mobile agents. Forudmobile agent-based systems to become more widely accepted there is a criticaludneed for a methodology to be developed to address various issues related toudmodelling mobility of agent . This research study provides an overview of theudcurrent approaches, methodologies and modelling languages that can be usedudfor developing multi-agent systems. The overview indicates extensive researchudon methodologies for modelling multi-agent systems and little on mobility inudmobile agent-based systems. An original contribution in this research known asudMobile agent-based Mobility Methodology (MaMM) is the methodology forudmodelling mobility in mobile agent-based systems using underlying principles ofudGenetic Algorithms (GA) with emphasis on fitness functions and geneticudrepresentation. Delphi study and case studies were employed in carrying out thisudresearch.