| With the rapid spread of mobile Internet, Internet of Things and cloud computing, the ossification of conventional Internet has become a prominent problem. Network virtualization promises to integrate current virtualization technologies of both communication and computing resources, and to overcome the ossification by a hierarchical layered structure. Among others, Software-Defined Networking(SDN) has provided a practical solution for the realization of such network virtualization. This thesis focuses on the system architecture and the resource allocation in the network virtualization based on SDN.The thesis firstly analyzes the characteristics of network virtualization and SDN technology, and discusses the architecture of network virtualization based on SDN. Then, the thesis elaborates the significance of virtual resource allocation which essentially is virtual network embedding(VNE). Furthermore, the thesis describes the VNE model in a mathematical form.Secondly, a resource allocation proposal based on rate adaptive energy savings is put forward for the purpose of the carrier-level deployment. In this proposal, the resource of substrate network is set to different levels by proportional computing technique. A algorithm is designed in order to solve the mixed integer programming problem. An open-source software ALEVIN is used to do the simulation experiment. Our results show that the algorithm has advantages in terms of energy savings and the virtual network request acceptance rate.Finally, this thesis proposes a resource allocation scheme aiming at load balancing. By introducing mathematical constraint variables, the nonlinear computational model is converted into a mixed integer programming model. The performance is evaluated numerically for both deterministic and randomized virtual network requests. Our results show that the proposed scheme is better in both load balancing and energy saving while keeping a high rate of service requests acceptance. |
PDF Full Text Request