Architectural components for the efficient design of mobile agent systems

  • Authors:

  • Marthie Schoeman;Elsabé Cloete

  • Affiliations:

  • Department of Computer Science and Information Systems, University of South Africa, PO Box 392, UNISA, 0003, South Africa;Department of Computer Science and Information Systems, University of South Africa, PO Box 392, UNISA, 0003, South Africa

  • Venue:
  • SAICSIT '03 Proceedings of the 2003 annual research conference of the South African institute of computer scientists and information technologists on Enablement through technology
  • Year:
  • 2003

Quantified Score

Hi-index 0.00

Visualization

Abstract

Over the past eighteen months, there has been a renewed interest in mobile agent technology due to the continued exponential growth of Internet applications, the establishment of open standards for these applications, as well as the semantic web developments. However, the lack of a standardised programming model addressing all aspects of mobile agent systems prevents widespread deployment of the potentially useful technology. The architectural requirements dealing with all aspects of a mobile agent system are not clearly stipulated. As a result, the commercially available mobile agent systems and mobile agent tool kits address different mobile agent issues, and little reuse of available technologies and architectures takes place. The purpose of this paper is to describe an architectural model that identifies the components representing the essential aspects of a mobile agent system. Due to the intensive nature of development, implementation and testing of this model, we describe preliminary work. However, in the meanwhile, there are benefits associated with this preliminary model, namely that it provides a clear understanding of the architectural issues of mobile agent computing, giving novice researchers and practitioners who enters the field for the first time a foundation for making sensible decisions when researching, designing and developing mobile agents. The model is also significant in that it provides a benchmark for researchers and developers to measure the capabilities of mobile agents created by commercially available tool kits.