 Hello. Good morning, ladies and gentlemen. My name is Yufang Hong. I'm from ZTE Corporation. Today, I'm very happy to be here to present this ZTE Common Edge solution, which is driven by your Cloud Cloud engine. As you may already know that in 5G, we have three main scenarios, which is EMBB, URLC, and MMTC. EMBB will provide 10 gigabits network bandwidth, and URLC will provide a guarantee 1 millisecond latency. MMTC will provide 1 million connections per square kilometers. So I mean, C is an important part to support these key features, which, especially for the URLC, without I mean, C is not possible to provide 1 millisecond latency. And also, for EMBB, we will need I mean, C to optimize the network traffic flow, which can reduce backhaul traffic load. And also, I mean, C will accelerate service innovation by providing a lot of rich capabilities of the network and the IT capabilities. So but still, we have challenges with I mean, C. Firstly, we have to think about how to introduce all our third-party applications. These third-party applications may need different running environments, like a bare metal virtual machine and containers. Also, some of the applications will need hardware acceleration to reduce the latency caused by the computing process. So this, like the ARVR, for the network transmission, for the network forwarding and graphic video processing, also certainly, how to deploy I mean, C easily and how to deploy I mean, C application easily in the place of the network edge, which has very limited results, like power space and personnel skills. So in the following slide, I'm going to show you how the common edge solution fulfills this challenge. Firstly, in this slide, I'm going to show you the overview of the common edge solution. Generally, we have four layers. One layer is an infrastructure layer, which is hardware composed by the different kind of servers. Either it's a commercial of the shell server or it's a specific server. With STIN defined, low latency network equipment. On top of that infrastructure layer, we have virtualization layer, which is light with open stack in combination with Kubernetes. We can provide the bare metal virtual machine and the machine in virtual machine. On top of that, on top of the virtualization layer, we have platform layer. This layer provides different capabilities. On the left part, we have TUF, Gateway, and UPF, which provides local traffic breakout functionality, which is used to reduce the network latency. On the right part, we have the network capability platform, which provides both city capabilities and IT capabilities, like the radio network information services, like location-based services, and TCP optimization. This belongs to the city domain. Also, we have the capability, like a video optimization and different IPP, management capability. Also, we have a lightweight ID manual and APVM to allow the simplified IPP deployment. On top of the platform layer, we have application layer. Here, we have different applications that may form different vendors. So in summary, we have focusing on the 4C, which is a connection, which provided by the UPF and different gateways. Also cloud, by the open stack and the Kubernetes. A compute, which have different computing results, all different hardware acceleration capability. And the capability belongs to IT and city domain. Then let's see how ZT provides a dual-cloud engine to enable the flexible resource scheduling. Here, normally, when we talk about virtual machine containers, we think about open stack and Kubernetes. Normally, we have open stack and Kubernetes separately with two separate result pool. But here, we work on the solution. It's a converged open stack and the Kubernetes. Here, we share a lot of components, like a keystone, like sender and Newton, so that we can create mixed cloud. It's one cloud. This cloud can provide different kind of resources. If you provide open stack and Kubernetes separately, which means it's two clouds. The hardware cannot be shared. Network has to be interconnected. It's not under one network. And certainly, on the hardware layer, this open stack and Kubernetes can share a computer and different storage, like the safe storage and disk race. Then we have research a lot to the hardware acceleration. We have invested a lot of human resources in the cyborg. Also, we have collaborated with Inter and Melox to work on the hardware acceleration from the smart card. And also, we work together with Emilia to support the GPU acceleration. Also, ZD, as far as the hardware is concerned, we have our own hardware acceleration card. Like, we have two different kind of hardware acceleration card, which we call SmartNIC. One is to support the GPU of loads, which is specifically used by the Gateway and the UPF. Also, we have a standard card which supports OBS overload, which can support the standard application to improve the network capability and reduce the time delay for the computing part. Then for the MEC deployment and application deployment, we have two different ways. Especially for the far edge, we would like to recommend to have an all-in-one machine, which can be delivered by an express company. And you can just plug and play. And this MEC automatically reduced in the configuration server, which is located in the network. And then the configuration server will distribute the configuration file. And automatically, the MEC will start the configuration. And then this MEC will work. On the other hand, for application deployment, we have two different ways. One is we use simplified AIVIL. AIVIL is a standard component, which is in every architecture. With simplified AIVIL, you can easily use our application to generate the VFD file so that you can easily import the IT application and deploy it by the AIVIL. But also, we provide the IT way to introduce the applications, which is provided by we call it application management, which is our product director. So that you can, with this component, you can subscribe or the resource from the cloud, which is like public cloud. You can subscribe to the cloud, and you can upload your application image. And you can start to deploy. You can select which location you would like to deploy. And then you can start the deployment. So this is nearly automatic. So here, I would like to make a small advertisement of the city of AIVIL. We have more than 500 applications in the global market. But also, we have more than 60 collaboration in the 5G domain. Also, we have more than 50 MEC applications in the market. Then I would like to show you some case study. This is what we have done with front telecom in Spain. And we have started the project by the end of 2018. And we have tried the VOLR, which is voice over error. And by the end of June, we have successfully make a demo of holographic video call. Here, you can see that with this holographic call, you can see a person, which is standing in front of you. And like a real person, but it's fake. The left person is a fake man. Also, we have a big project together with Channel Mobile, which is in Guangzhou branch. And as you know, Channel Mobile is a very big operator. Here, we have four different MEC, which has been deployed for different locations. The access room, the aggregate room, and the call aggregate and the call room. We have deployed together with our partner for different kinds of applications, which like here, in the access room, we have Hongda Park. We're together with Hongda Park, and we try the V2X application. And together with the IMAQ and the network slating function, we provide a low latency capability. On the aggregate room, we work together with Tencent. We provide high definition content, stadium, and cloud gaming. Also, we have worked with Handtaly with video broadcasting, like live TV services. And in Jiangmen call room, we have worked with Aspire. We have provided facial recognition function. This is the first IMAQ deployment with dual call, which is Kubernetes and OpenStack dual engine, and hardware acceleration with DT OBS and GDB acceleration card. This is the latest reward we got from the global industry. First of all, the 5G excellence, which we just received from 2019 SDN World Congress in Netherland. And the greatest commercial potential for IH computing concept awarding the 2019 IH computing Congress in London. OK, here is a summary for your takeaway. In general, the ZD Common IH solution provides three main features, which I presented today. The powerful core engine, which we brought the OpenStack features into the Kubernetes so that they can share different components and provide unified cloud. We can provide a bare matter of machine and continuous simultaneously. Also, we can support the heterogeneous acceleration provided by ZT acceleration card, hardware, or third-party hardware. And also, remember that MEC in foreign has to be easy to deploy and for the application, it has to be easy to deploy. And also, we need a centralized management in a central location. We cannot have those good people in a far edge. OK, this is all my presentation. Any questions? Thank you.