A Trust-Aware, Self-Organizing System for large-scale federations of Utility Computing infrastructures

On-demand computing environments, like Cloud/Grid systems, consist of nodes that individually manage local resources which can be provided to users. When a client asks for some resources, it has the problem of finding the most suitable nodes satisfying such a request. In addition, the emerging context of large-scale federation of computing infrastructures is highly competitive, i.e., a node is in competition with each other to obtain the assignment of available tasks, such that the trustworthiness of nodes, which declare their capabilities, has to be taken into account. Indeed, Federated nodes, as in competition, in order to gain an advantage over its competitors, or due to different factors of inaccuracy, may publish a wider set of resources/services than those currently available for it. Once the task has been assigned, these nodes, to complete the task, could require the collaboration of other nodes (or they would be unable to perform the job due to the real lack of resources), hence the problem becomes to find the most promising collaborators. In such a competitive scenario, to make an effective selection of possible collaborators, each node should be provided with a trust model for accurately evaluating the trustworthiness of its interlocutors. In this paper a trust-based approach for large-scale federations Utility Computing infrastructures is proposed. The discussed model is designed to allow any node to find the most suitable collaborators in an efficient way, by avoiding to explore the whole node space. The presented approach exploits a fully decentralized approach which allows to organize the nodes of the federation in an overlay network which eventually exploits some specific properties, such that each customer and provider looking for collaborators will be capable of determining the set of nodes to be considered in searching for collaborators.