Edge Network Task Scheduling and NFV Deployment Strategies Toward Future 5G Service On-Demand Applications

• Main Authors: Chia-Wei Tseng, 曾家偉
• Other Authors: Li-Der Chou
• Format: Others
• Language: en_US
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/262un3

博士 === 國立中央大學 === 資訊工程學系 === 107 === With the rapid development of mobile, cloud and Internet of Things (IoT), the demand for network resource allocation, traffic processing and service management drives the transformation of traditional network infrastructure. The emergence of Software Defined Network (SDN) and Network Functions Virtualization (NFV) technology turns the complicated network architecture into a virtual and programmable network. SDN/NFV not only drives the transformation of the Information and Communication Technology (ICT) industry, but also the rise of edge computing, leading the development trend of 5G technology in the future. How to meet the requirements of different users, flexible and rapid allocation of virtual computing resources, and further provide on-demand service is the key to the future development of 5G IoT services. In order to reduce the waiting time of data to and from the cloud and reduce the network bandwidth cost, this paper proposes a gateway-based edge computing service model. By adjusting the edge gateway's task schedule, more service requests can be processed with less waiting time. The user's service requests can be processed as close as possible to the edge network devices. If the computing power required by the service exceeds the computing power of the edge gateway or cannot be processed, it will be forwarded to the cloud for processing. In terms of Virtual Network Function (VNF) deployment, the application of NFV virtualization technology and SDN OpenFlow traffic control mechanism can significantly improve the flexibility and scalability of network service deployment. In order to achieve rapid deployment of NFV, this paper analyzes several different deployment strategies and explores factors that may affect the efficiency of VNF deployment. In addition, this paper also compares the differences between Virtual Machine (VM) and Docker virtualization technologies to meet the requirements of edge computing device resource allocation. In terms of edge computing applications, because network security is an important research direction of 5G, the paper utilizes Service Function Chaining (SFC) technology to design and implementation of a Service On-Demand (SOD) system for security applications. SFC can improve the application efficiency of network security services and reduce the cost of hardware equipment and provide innovative action edge network security application and service models that can meet the needs of different users. The edge computing service architecture proposed in this paper helps to reduce the computational load of traditional cloud architecture and improve the operational efficiency of edge computing devices. It can be used as the basis for the development of Cloudy-Fog Computing integrated security applications in the future 5G network.