Fuzzy cognitive maps for dynamic grid service negotiation

The grid is moving from the scientific grid to a pervasive and economic/business grid. Service trading, in which service provider and service consumer negotiate for a mutually acceptable agreement on multi-issues such as service performance, access cost etc., is one of the most important components in building the Economic Grid. In view of Pervasive Grid, a new challenging issue is that participants on pervasive devices usually have limited computational capacity. And it is also desirable that a multi-issue negotiation agreement can be reached as quickly as possible since the wireless communication to exchange the offers is generally unreliable and power-consuming. Hence, an agile, automated, but lightweight multi-issue decision-making model is needed to facilitate service negotiation in Pervasive Grid. More over, existing methods for multi-issue negotiation only regard each issue as a separate issue, though in most of cases, there exist causal relationships between these negotiation issues. In this paper, a decision-making model based on Fuzzy Cognitive Map (FCM) theory is proposed for multi-issue negotiation which takes into account the causal relationships between the negotiation issues. In the proposed model, the causal relationships between the negotiation issues are well represented by FCMs. The service trading is modeled as a dynamic system with interdependent relationships among negotiation issues. The example and experimental results show that the proposed model is lightweight and promising to be employed in Pervasive Grid for multi-issue service negotiations.