In the 5G era, what kind of data center do operators need?

The evolution of mobile communication networks is about one generation every ten years. With the arrival of 2018, the general public is one step closer to the 5G era. However, behind the scenes, an invisible "war" around the layout of the 5G era has already quietly started among operators.

In the 5G era, operators are standing at a new starting line

When it comes to 5G, the first thing that comes to mind is probably its high speed and high bandwidth features. For example, it takes 1 second to download a movie, watch all kinds of high-definition videos through mobile terminals, and transfer large files. In fact, 5G brings more than just an increase in speed.

As a new generation of mobile communication technology, 5G has the characteristics of high speed, low latency, low power consumption, high capacity, and wide coverage, and is known as "a technology born from the Internet of Things". It can be foreseen that 5G will greatly eliminate the connection barriers between people, things, and people and things, and strongly promote the development of the Internet of Things. At the same time, it will also trigger tremendous changes in the telecommunications industry itself and all walks of life in society.

It is worth noting that new technologies are bound to generate new application values ​​and new market opportunities. According to Gartner's forecast, by 2020, 5G will connect 50 billion smart devices and 7.7 billion people. By then, 5G will promote a large demand for video content services, and will also generate new application demands such as AR, VR, driverless driving, Internet of Vehicles, and smart homes. These also mean broad market opportunities for operators.

[[220343]]

IT platform capabilities are key

There is no doubt that with the advent of the 5G era, challenges are also obvious for operators. In my opinion, 5G will challenge the various capabilities of operators' cloud platforms, including the following aspects:

First, due to the breakthrough of data transmission bottlenecks such as bandwidth and latency, smart terminals and IoT devices are no longer limited by the computing power of the devices themselves, but can leverage the cloud to obtain more powerful data processing capabilities. Therefore, it poses a challenge to the platform service and support capabilities of operators' cloud data centers.

Secondly, with the rapid development of new services such as high-definition mobile video, AR, VR, driverless, Internet of Vehicles, and smart home, the network is required to provide ultra-low latency, localized service processing, customization, and other specific capabilities. This requires sinking some functions of the mobile core network to be closer to the content source.

For example, advanced technologies such as driverless cars have very high requirements for real-time data such as vehicle operation status and road conditions, and they also need to maintain frequent interactions with data centers. Therefore, these new requirements will bring challenges to the capabilities of operators' cloud data centers.

***, 5G technology will also stimulate a large number of business innovations. The increase in the types of operators' services will also increase the complexity of technology and the difficulty of management. This requires operators to have agile and flexible application deployment while also having strong automation and management and operation capabilities.

It is not difficult to see that the real challenge for telecom operators with the advent of the 5G era comes mostly from the back-end platform. So, how should telecom operators build cloud data centers for the 5G era? In Huawei's view, telecom operators need to focus on the following areas:

***, edge computing. As mentioned above, in the 5G era, IoT services such as AR, VR, and V2X have extremely high requirements for network latency and speed. Not only do they need to move network functions and service processing functions down to locations close to the access network, but they also place higher requirements on computing and forwarding capabilities. Edge computing can undoubtedly meet these requirements well.

Second, smart engines. In the 5G era, there will be image and video recognition in wireless video surveillance scenarios, analysis of vehicle GPS location trajectories and driving behavior preferences in Internet of Vehicles scenarios, congestion and violation detection, and traffic light optimization in smart traffic scenarios, group density and mobility pattern prediction in smart city scenarios, and power usage distribution and peak prediction in smart grid scenarios. The above-mentioned "smart engines" for 5G IoT scenarios require cloud data centers to rely on the IoT data lake gathered by the big data platform to further provide rich and complete basic platform services that pre-integrate the industry's mainstream machine learning, deep learning, graph engines, search algorithms, and other basic platform services.

Third, more efficient hardware form. 5G requires a rate and bandwidth improvement of more than 100 times compared to 4G networks, and also puts forward more stringent requirements on reliability and latency guarantee, in order to meet the needs of typical industry application scenarios such as virtual reality, ultra-high-definition video, intelligent manufacturing, and autonomous driving in the future. Therefore, under the premise of strictly complying with the open service API and ecology of the industry's de facto standards, whether the high network throughput, high concurrent computing and storage IO requirements of the cloud platform for 5G networks and services can be met, and the overall performance and cost of the cloud data center can be optimized, has become a touchstone for measuring the competitiveness of cloud data centers in the 5G era.

5G is here, and operators need to accelerate their technology deployment to seize the opportunity.