Resource Allocation in Market-Oriented Clouds

In mid-1990s, the vast deployment of high-speed public data networks has motivated the externalization of private computing resources. Cloud is the most recent evolution of distributed computing. In terms of resource virtualization, managing conjunctly distributed computing, storage, and network facilities remains a complex optimization problem. This problem is closely related to the design of innovative pricing strategies for Cloud Service Providers (CSPs) that act as mediators between resource owners and end-users. In practice, resource owners do not accept, for confidentiality reasons, to inform the CSP about a detailed description of their available equipment. Meanwhile, they accept to regularly provide the CSP with an abstracted view of their sharable resources. In this book, we propose market-based resource allocation algorithms. We consider two business models depending on the fact the CSP owns or not the required resources. An extension to our algorithm is to make use of multicast nodes in order to manage point-to- multipoint data transfer. Another feature is to distribute storage data over multiple storage nodes.