5G is the first generation of cellular born in the cloud.
Like electricity and water, cellular connectivity has become a commodity that’s an essential part of modern society, the medium that will allow innovation in real-time data environments. and its ability to innovate in real-time data environments. Increasingly mobile and distributed workforces will collaborate in new, more meaningful ways; industries will cut the wire, gaining flexibility and data insights that reshape products and process; the automation of daily tasks will give us all more time to work, to play, to connect. And as service providers and their attendant ecosystem of vendors embark on the journey of commercializing, scaling and maturing 5G, the entire industry must begin provisioning wireless connectivity as a commodity.
Enhanced mobile broadband, massive Internet of Things and ultra reliable low latency communications for mission critical services are the three primary 5G use cases.. Successfully delivering 5G applications requires growing the compute power present in traditional data centers as well as at the network edge. In a previous post, we explored how and why cloud is a crucial piece of 5G; now we’ll examine some of the best practices service providers should observe in order to keep pace with a rapidly changing technological and business environment.
In an interview with RCR Wireless News, Dell EMC Global CTO John Roese explained how the company is concurrently working with service provider customers to develop the appropriate network architecture, topology and cloud strategy to support 5G and engaging with enterprise customers to explore how 5G will impact their business, “acting as a broker of sorts.”
“On the telecom side, we’re really in the phase right now where the operators are making sure they build the right architecture. 5G is not just an evolution or a faster 4G–it’s a whole new concept and it’s as much of an IT project as a telecom effort. With our enterprise base, we have these discussions to understand what their unmet needs are. We’re able to identify potential new areas for us to collaborate on and that results in new use cases, which we can take to our telecom partners. Then we can go back to enterprise customers to help them develop their timing. They don’t want to jump too fast. Timing is everything.”
Extending the cloud from the data center to the edge
Ubiquitous applications like cloud-based photo storage, social media and streaming video services rely on large, centralized data centers packed with servers loaded with data storage and processing power. But as 5G takes shape, a new breed of applications that take advantage of single-digit millisecond latency will become possible and pervasive to meet this data demand. Mobile virtual reality will let contractors see what a finished building should look like as it’s still under construction; high-quality telepresence will let doctors diagnose patients still en route to the hospital; and industrial robotics will gain real-time responsiveness and a new level of reliability.
To make all of that possible, the functionality associated with centralized data centers will have to move closer to the network edge–whether that’s a device in a users’ hands, an enterprise local area network or a cellular-connected machine or vehicle. Based on current market conditions, the build-out of edge compute infrastructure will take several forms dependent on the application. Operators are adding compute power to existing cellular sites–macro towers and colocation facilities, and third-party providers are standing up edge data centers in proximity to existing fiber-optic lines. In either case, the common denominator is IT infrastructure capable of dynamically supporting a wide variety of workloads that can be flexibly tailored to fit the desired application.
Virtualization beyond the core
The massive increase in cellular data consumption, which will only continue to increase, requires operators to scale up through investment in complex (and expensive) network infrastructure, which isn’t sustainable. Given this paradigm, operators have switched gears and, to varying degrees, adopted a more IT-focused approach by using network functions virtualization and software-defined networking. Single-purpose hardware is being replaced with off-the-shelf hardware capable of recreating hardware-dependent functions virtually.
In the core network, where the intricacies of network operations take place, NFV and SDN allow for the separation of the network and control planes. This enables the abstraction of formerly complex, manual processes into software. Now the network can be reconfigured more quickly through automated processes, making the network more responsive to user needs more quickly, which lets operators bring new services to market much more quickly, creating new revenue opportunities while delivering a better experience to subscribers.
While it’s still relatively nascent, the next step for virtualization is extending the concept beyond the core and into the radio access network. Instead of creating an RF signal from a proprietary baseband unit, the same thing can be accomplished more quickly and more cost effectively through the use of commercial-off-the-shelf IT infrastructure. In the context of 5G, which will require a mix of low-, mid- and high-band frequencies to support the wide array of existing and future applications, this IT-centric approach again saves service providers time and money. Software control can automatically pair network and spectral resources with the service-level requirements for a particular application.
Automating the network
As PCs begin shipping with embedded cellular connectivity, joining the billions of smartphones already dependent on global networks, coupled with exponential growth in the connected devices comprising the Internet of Things, managing future 5G networks will only be possible with automation.
Centralized and distributed workloads, ranging from device authentication and security, signal processing and routing, customer service and billing, and storing, analyzing and monetizing massive amounts of data, will be seamlessly managed by artificial intelligence-powered software running on a dazzling array of general purpose hardware. If this is the destination, it’s clear that the journey to 5G will be long, fraught and best navigated with a clear roadmap and trusted partner.
Looking to this automated future, Roese emphasized the need to essentially create time. “The reason we have to create time is because don’t have enough human capacity to meet the demand. If we don’t build high degrees of automation everywhere, we won’t be able to get to our outcomes. The existential issue is if we don’t do this across all of our IT systems, we’ll run out of people and industries will stagnate.”
To learn more click here.
