5G Is About More Than Just Smartphones
Wireless carriers have begun rolling out 5G network upgrades in different parts of the U.S. Indeed, by the end of 2019, Verizon plans to have 5G available in thirty cities. With 5G connections, users will recognize higher speeds, but 5G brings far more than faster smartphone connections. When people talk about 5G, they are really talking about a set of technologies that will enable new applications and innovations that go far beyond the smartphone.
The 5G Revolution
5G isn’t just about having faster smartphones. 5G is designed to be the foundation of a broader network ecosystem. The increased availability of efficient data connections for small devices will advance the Internet of Things (IoT)—in fact, this use case is so important that DISH is building an IoT-specific network. 5G connectivity will enable massive machine-type communications—that is, seamlessly connecting embedded sensors in virtually anything. This will play a critical role in industrial IoT, enabling smart cities, smart utilities, and security infrastructure.
Providing intelligence at the edge of the network will lower latency and distribute processing, similar to the way that content distribution networks, like Cloudflare’s, have enabled new applications on the traditional Internet. These innovations will enable new 5G use cases that rely on low-latency and distributed processing such as augmented reality (AR) and virtual reality (VR), robotics, automated vehicles, advanced manufacturing, and healthcare imaging.
The arrival of commercial 5G networks will create a new network topology that will further encourage innovation in many areas other than smartphones. The revolutionary nature of 5G is enabled by a number of innovative aspects of 5G architecture, especially network function virtualization, edge computing, and network slicing.
Network Function Virtualization
Network function virtualization (NFV) refers to the process of decoupling networking functions from dedicated hardware into software. In LTE networks, a service provider deploys network functions by having a specialized appliance for each individual function. A different piece of hardware was required to complete each task—a separate appliance for the router, firewall, encryption function, and so on. To expand the network, a customer has to acquire more specialized hardware, and if the network needs more encryption but less firewall capacity, some equipment will go unused. This makes it harder to scale the network up.
Source: TelecomTutorial info: Introduction to NFV Network function Virtualization Basics, YouTube, https://www.youtube.com/watch?v=Vl5UJUR1uV4.
NFV replaces the expensive hardware with a generic server that could perform any network task. Software packages allow generic computer hardware to deploy multiple network functions simultaneously, decoupling functionality from computing capacity. This enables service providers to scale capacity up or down very easily.
This process of migrating the network core and edge processing from the physical to the virtual, in preparation for 5G, took Verizon three years. Adam Koeppe, Senior Vice President of network planning at Verizon, boasted that the 5G network cores that it launched in the four 5G Home markets are 100 percent virtual. “Unlike LTE where you had to start physical and move to virtual, they’re native 5G network functions, those all start as virtual,” Koeppe said.
Edge computing has also been key to building out 5G. This refers to generating, collecting, and analyzing data on the “edge” of the network in a cloud environment. Network function virtualization allows the same machines that provide the processing power for network functions to also process data for customer applications. By storing and processing data at the edge of the network, latency for the end user can be lowered significantly. Services enabled by 5G will require low latency—the amount of time that it takes a message to traverse a computer network system. Applications like AI, game streaming, and augmented reality can require latency of no more than 20 milliseconds for data to be processed close to the end user. Safety-critical applications like industrial automation control and autonomous driving may require latencies in the single millisecond range. A recent trial on an Intel/AT&T 5G test network delivered sub-10 millisecond latencies, with lower latencies expected in production networks.
Finally, 5G network slicing of the 5G architecture will allow operators to provide portions of their networks for specific customer use cases—whether that be a smart home, autonomous vehicles, or a smart energy grid. Network slicing is a type of virtual networking architecture in the same family as NFV. By building multiple virtual networks atop a shared physical infrastructure, network slicing allows for optimized resources that suit the needs of an application. Each “slice” or self-contained partition shares common storage and processors with the other partitions. Software-based management devotes capacity to each partition dynamically according to need.
Source: What Is 5G Network Slicing, SDxCentral, https://www.sdxcentral.com/5g/definitions/5g-network-slicing/.
Some companies, like Ericsson, see network slicing as the key ingredient for 5G meeting technical requirements for customers. The 5G connectivity will allow for a diversity of use cases, and the elasticity brought by network slicing will help accommodate the varied requirements in terms of power, bandwidth, and speed.
To ensure U.S. competitiveness across the 5G revolution, it’s important to protect contributors throughout the production chain, not just the narrow segment of companies that make smartphone modems. Policymakers should consider the companies that make smart devices and the 5G modules that will provide them with connectivity, as well as the companies making the hardware and software that will enable edge computing and network virtualization. Each of those steps in the 5G value chain provides new challenges and new benefits, but privileging any one aspect of 5G over others risks winning the battle but losing the war.