Research On Networking And Application Of Mobile Edge Computing

As the number of mobile devices has increased dramatically,Mobile Internet, Internet of Things (IoT), and Mobile Cloud Computing (MCC) are driving a new round of information and communication industry integration. MCC has brought a huge space for the development of Mobile Internet and new business chain.However, it is difficult to meet the requirements for low-power,large-connected, high reliability and low latency of 5G by the way that traditional MCC used to upload data processing, intelligent computing to the remote cloud.Mobile Edge Computing (MEC) provides mobile and cloud computing capabilities within the access network, and aims to unite the telco and IT at the mobile network edge. It is a new network architecture concept that offers a service environment with ultralow latency, high-bandwidth, and direct access to real-time network information.In order to meet the requirements of low latency, high bandwidth and network openness, this paper mainly focused on the mobile edge computing networking model to realize the convergence of access network and cloud computing. What's more, we deployed a MEC platform based on the proposed MEC networking model, and developed a series of location-aware applications to verify its performance and openness of mobile edge computing. This paper mainly completed the following work:(1) Researched the architecture of MEC and the traditional MEC networking scheme. By analyzing the typical MEC use cases,concluded that the key issue to be solved in MEC networking is the effective integration and flexible scheduling of computing, storage and access network resources. To solve the problem, this paper proposed a MEC networking model based on the convergence of access network and edge cloud center. In our model,software-defined network (SDN) is used to build the access network,and SDN controller grantees the communications between and within the edge data center constituted by MEC servers. Additionally,through the Network Function Virtualization (NFV) scheduling mechanism, edge cloud center can allocate and manage the vitual computing, storage and networking resources flexibly.(2) According to the proposed MEC networking model, we deployed a mobile edge computing platform-WiCloud, which is constituded by industrial servers, OpenFlow switches and 802.11ac access points (APs). WiCloud uses SDN controller to build the network of WiCloud and cloud services are deployed in virtual machines hosted on it. Meanwhile, we developed an information moduler to realize the function of data acquisition, and it can share the information of subscribers' mobile devices around with applications running on WiCloud.(3) We deployed 40 WiFi APs in the campus of Beijing University of Posts and Telecommunications (BUPT), and developed a series of location-aware applications including analysis of the spatial distribution of campus population characteristics, quality of campus wireless network and time distribution characteristics of crowds' traffic flow. Additionally, we verified the performance of the MEC platform by watching 4K video in a MEC environment compared with that in MCC. The result shows that an open MEC ecosystem can transfer the real-time radio network information into innovative applications, and the charactersize of low latency and high bandwidth.Mobile Edge Computing can migrate cloud computing to the access network and provide a low latency and high bandwidth application environment for mobile applications and mobile services.This paper studies the networking and application of mobile edge computing, and verified that MEC has a big advantage in delay,bandwidth and other network performance over mobile cloud computing. What's more, the opneness of MEC can promote the development of innovative mobile applications and cloud services.