A proposal of survivable virtual network embedding algorithm

Abstract

People heavily depend on Internet in their daily lives and propose diversified requirements, such as security and privacy. Network virtualization is considered a promising approach to overcome the disadvantages of the current networks. Therefore, how to reduce the node or link failures resulting fr om network outages and achieve high-quality network services should be the goal for Internet service providers to improve network security and quality of service. For network survivability, a failure protection mechanism that reserves secondary resources before any failure occurs is presented in this paper. In this study, network link resources are divided into primary and secondary flow resources. Primary flow resources are used under normal network operation, whereas secondary flow resources are used as backup resources when the network fails. When a VN request arrives, a survivable VN embedding algorithm based on a group search optimizer (GSVNE) is employed to search for optimal embedding solutions. The resource allocation proportions of the primary and secondary flow resources can be dynamically adjusted. Simulation results show that GSVNE outperforms the conventional greedy algorithm, which reflects desirable results in authoritative papers, in terms of cost and acceptance rate.