Challenger: a multi-agent system for distributed resource allocation
AGENTS '97 Proceedings of the first international conference on Autonomous agents
Connection establishment protocol based on mutual selection by users and network providers
Proceedings of the first international conference on Information and computation economies
Market-driven service allocation in a QoS-capable environment
Proceedings of the first international conference on Information and computation economies
Market based bandwidth allocation policies for QoS control in broadband networks
Proceedings of the first international conference on Information and computation economies
Dynamic behavior of differential pricing and quality of service options for the Internet
Proceedings of the first international conference on Information and computation economies
The POPCORN market—an online market for computational resources
Proceedings of the first international conference on Information and computation economies
On tariffs, policing and admission control for multiservice networks
Operations Research Letters
Hi-index | 0.24 |
A reservation protocol and pricing model is proposed for the allocation of electronic services, such as computing and communication resources. Whereas centralised allocation mechanisms are frequently analysed, and therefore better understood, we focus on a decentralised, bilateral market, employing a new protocol to enable application agents to request price quotes and to book resource capacity from resource providers. In contrast to existing protocols we allow for competition between providers. Contracts and prices are confidential; therefore, resource providers can only collect information from their own bilateral trades and from requests for price quotes. They determine the optimal price with a supply function that is extended to continuous trading by updating the optimal price quotes with new information. We implement the protocol in a simulation, show that the equilibrium prices are above competitive levels and demonstrate that simple deviations from the pricing model are not profitable.