A Real-time Group Auction System for Efficient Allocation of Cloud Internet Applications

摘要

Increasing number of the cloud-based Internet applications demands forefficient resource and cost management. This paper proposes a real-time groupauction system for the cloud instance market. The system is designed based on acombinatorial double auction, and its applicability and effectiveness areevaluated in terms of resource efficiency and monetary benefits to auctionparticipants (e.g., cloud users and providers). The proposed auction systemassists them to decide when and how providers allocate their resources to whichusers. Furthermore, we propose a distributed algorithm using a group formationgame that determines which users and providers will trade resources by theircooperative decisions. To find how to allocate the resources, the utilityoptimization problem is formulated as a binary integer programming problem, andthe nearly optimal solution is obtained by a heuristic algorithm with quadratictime complexity. In comparison studies, the proposed real-time group auctionsystem with cooperation outperforms an individual auction in terms of theresource efficiency (e.g., the request acceptance rate for users and resourceutilization for providers) and monetary benefits (e.g., average payments forusers and total profits for providers).