Scalability and robustness of a network resource allocation system using market-based agents

Authors:
Nadim Haque;Nicholas R. Jennings;Luc Moreau
Affiliations:
School of Electronics and Computer Science, University of Southampton, Southampton, UK SO17 1BJ;School of Electronics and Computer Science, University of Southampton, Southampton, UK SO17 1BJ;School of Electronics and Computer Science, University of Southampton, Southampton, UK SO17 1BJ
Venue:
Netnomics
Year:
2005

Citing 0
Cited 3

An Auction-Based System for Workforce Resource Allocation

IEA/AIE '08 Proceedings of the 21st international conference on Industrial, Engineering and Other Applications of Applied Intelligent Systems: New Frontiers in Applied Artificial Intelligence
Evolutionary Market Agents for Resource Allocation in Decentralised Systems

Proceedings of the 10th international conference on Parallel Problem Solving from Nature: PPSN X
Evolutionary market agents and heterogeneous service providers: achieving desired resource allocations

CEC'09 Proceedings of the Eleventh conference on Congress on Evolutionary Computation

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper, we consider issues associated with scalability and robustness in designing a market-based multi-agent system that allocates bandwidth in a communications network. Specifically, an empirical evaluation is carried out to assess the system performance under a variety of design configurations in order to provide an insight into network deployment issues. This extends our previous work in which we developed an application that makes use of market-based software agents that compete in decentralised marketplaces to buy and sell bandwidth resources in a network that is partitioned into regions, each with a separate market server. We investigate the average call success rate and average message load per market server, as the number of markets are scaled up in a fixed size network. The same investigations are performed in the presence of single market failures. Finally, for both the failure and non-failure cases, a trade-off is found between their average call success rates and message load per server in order to find an optimum number of regions to deploy in the network.