Construction and Resource Allocation of Cost-Efficient Clustered Virtual Network in Software Defined Networks

As cloud computing becomes more and more massive and dynamic, software defined networks (SDNs), as a global sight of the network, make resource allocation more serviceable and effective. However, it is a great challenge to effectively allocate resource for different tenants’ computing requests while considering the performance of the whole network. To address this problem, in this paper, we propose a clustered virtual network (CVN) abstraction strategy and a dynamical heapsort algorithm to find the pod with the most appropriate number of virtual machines. In the cloud, when the SDN controller receives a tenant’s computing task request, it will search for the resources in the whole data center resource pool by the tenant’s request. If there are enough virtual machine resources in one single pod to meet the computing task, the controller will construct a CVN time-efficiently. If not, the controller will provide resources from several pods to construct a crossing pod CVN. Furthermore, we formulate the resource allocation problem as a Linear Programming problem aiming to maximize network throughput. To achieve computation feasibility in massive data center networks, we propose an approximation primal dual algorithm for solving this Linear Programming problem. Theoretical analysis shows that the proposed primal-dual algorithm is feasible and suitable for solving massive computation problem in SDNs. Finally, we evaluate the performance of our proposed algorithms and verify the correctness of our theoretical analysis.