A resource occupation ratio based virtual network balanced mapping algorithm in cloud computing environment

Diverse services need different resources. It is an important issue to achieve the resources occupied by services evenly distribution in network, i.e., avoiding this situation that some devices are fully or even over loaded, while others are free. To solve this problem, in cloud computing environment, the redundant paths in network and data flow partition technology are exploited. The virtual resource occupied by service is modeled, introducing node stress index and link stress index to evaluate the distribution of service occupied network resource, which is abstracted as a non-linear programming problem. A virtual network balanced mapping algorithm is proposed to evenly distribute occupied network resources and improve resource utilization ratio. At last, through numerical simulation, proposed algorithm is superior to existed work in the terms of sub-path number, accepted service request number, etc.