Scalable mechanism design for the procurement of services with uncertain durations

  • Authors:

  • Enrico Gerding;Sebastian Stein;Kate Larson;Alex Rogers;Nicholas R. Jennings

  • Affiliations:

  • University of Southampton, Southampton, UK;University of Southampton, Southampton, UK;University of Waterloo, Waterloo, ON, Canada;University of Southampton, Southampton, UK;University of Southampton, Southampton, UK

  • Venue:
  • Proceedings of the 9th International Conference on Autonomous Agents and Multiagent Systems: volume 1 - Volume 1
  • Year:
  • 2010

Quantified Score

Hi-index 0.01

Visualization

Abstract

In this paper, we study a service procurement problem with uncertainty as to whether service providers are capable of completing a given task within a specifie deadline. This type of setting is often encountered in large and dynamic multi-agent systems, such as computational Grids or clouds. To effectively deal with this uncertainty, the consumer may dynamically and redundantly procure multiple services over time, in order to increase the probability of success, while at the same time balancing this with the additional procurement costs. However, in order to do this optimally, the consumer requires information about the providers' costs and their success probabilities over time. This information is typically held privately by the providers and they may have incentives to misreport this, so as to increase their own profits To address this problem, we introduce a novel mechanism that incentivises self-interested providers to reveal their true costs and capabilities, and we show that this mechanism is ex-post incentive compatible, efficien and individually rational. However, for these properties to hold, it generally needs to compute the optimal solution, which can be intractable in large settings. Therefore, we show how we can generate approximate solutions while maintaining the economic properties of the mechanism. This approximation admits a polynomial-time solution that can be computed in seconds even for hundreds of providers, and we demonstrate empirically that it performs as well as the optimal in typical scenarios. In particularly challenging settings, we show that it still achieves 97% or more of the optimal.