Low-cost itineraries for multi-hop agents designed for scalable monitoring of multiple subnets

  • Authors:

  • Damianos Gavalas;Christina Tanya Politi

  • Affiliations:

  • Department of Cultural Technology and Communication, University of the Aegean, Mytilini, Lesvos Island, Greece;School of Electrical and Computer Engineering, National Technical University of Athens, Zographou, Athens, Greece

  • Venue:
  • Computer Networks: The International Journal of Computer and Telecommunications Networking
  • Year:
  • 2006

Quantified Score

Hi-index 0.00

Visualization

Abstract

Mobile agent (MA) technology represents a recent trend for implementing distributed management architectures, as an answer to the flexibility and scalability problems of centralised models. Management scalability limitations though, are not adequately addressed when monitoring tasks requiring the employment of multi-hop agents are considered. This is because agent-based architectures lack mechanisms that guarantee near-optimal agents itineraries so as to minimise the total migration cost in terms of the round-trip latency and the incurred traffic. This is of particular importance when the management of networks spanning multiple subnets is involved. To address these issues, we have adapted an algorithm originally designed to solve network design problems to the specific problematic of MA itinerary planning. The algorithm suggests the optimal number of MAs that minimise the overall cost and also constructs optimal itineraries for each of them. A Java-based implementation of the algorithm has been tested on realistic applications over a laboratory testbed, demonstrating significant cost savings. Simulation tests verified the algorithm's validity and competence over large enterprise networks.