Modeling and Delay Analysis for SDN-based 5G Edge Clouds

Abstract

The fifth generation (5G) mobile networks are envisioned to provide connectivity not only to mobile users but also to a wide range of other services such as enhanced mobile broadband (eMBB) and massive Internet of Things (mIoT). In order to meet the diverse requirements of these services in 5G, Software Defined Networking (SDN) has been proposed as an enabling technology for both the core cloud and the edge cloud, in addition to Network Slicing to achieve isolation among services. In this paper, an analytical model is developed for such an SDN-based edge cloud, focusing on the support of two services: eMBB and mIoT. To illustrate the use of the model, delay analysis of a switching node in the edge cloud is presented. The results show the relation between the packet delay and the underlying system parameters, such as slice density, and the impact of the SDN controller on the delay. An implication of the model, analysis and results is that they may be used for network / resource planning and admission control in 5G edge clouds to meet delay requirements of the services.