霧端無線接取網路之虛擬機配置設計之最小化回程網路流量

Abstract

With the advance of wireless technologies, the rapid mobile data traffic growth contributed by real-time interactive services leads to severe bandwidth consumption much more than the backhaul network capacity. Fog radio access networks (F-RAN) with virtualization technologies are one of the promising solutions used to reduce the backhaul bandwidth consumption in nextgeneration cellular networks. However, the placement of virtual machines (VMs) in F-RAN will significantly affect the backhaul bandwidth consumption and has to consider several issues. This thesis studies the replicationbased VM placement problem in a long-term aspect under multiple resource constraints over F-RAN and considers that each VM can serve multiple users. The objective is to minimize the total backhaul traffic. Then, we propose a heuristic algorithm to solve the target problem. Moreover, we observe that the VM re-allocation should not be frequently executed to actually minimize the total traffic generated in backhaul network. The simulation results agree our observation and show comparison with a greedy approach and the optimal solution, and the proposed algorithm demonstrates with favorable results for the overall backhaul network usage.