If one pays attention to the hype, 5G and edge networks are the solution to everything. Throw in a bit of AI, NFV, SDN, DSS, MIMO, mmWave, etc., and the wireless world will assume a life of its own, sans any overwatch. They might even self-deploy one day.
Seriously, however, between the edge and 5G, that is not that far from the truth once both get real traction. In fact, the edge, coupled with wide-bandwidth spectrum (read: mmWave), will handle the majority of data communications – and autonomously. However, do not look for this to resemble any kind of reality before about 2025.
I am not going to go into a lot of detail as to why I believe that will be the case. There is plenty of analysis and assessment of both platforms if one is interested in drilling down on the state of these and what other analysts and SMEs think.
It is generally agreed that the majority of 5G-based edge networks will be in highly populated areas running on the mmWave spectrum. The variety of edge cells is virtually unlimited – campuses, malls, city centers, ad hoc emergency communications, venues of all types, the enterprise space, and consumer locations to name a few. There will also be edge networks appearing in manufacturing, transportation, agriculture, mining, and other infrastructures, and, of course, the Internet of Anything/Everything (IoX) (I did not mention telecom because that is a given). With all of these applications (and their subsets), one can see why there is thought that 5G and the edge will be the solution for ubiquitous communications for the majority of applications for nearly everything.
So, why is this such a promising platform? Let us drill down on a few major vectors that can shed some light on this.
Of course, the two most enabling and hottest 5G technologies are low latency and speed (bandwidth). Those are a given. They are the core technologies that will enable the benefits and applications. Dynamic spectrum management as well as frequency agility come in a close second.
The architecture of edge networks – the distributed model of computing – is why all of this works. Couple that with the components just mentioned, and the resultant networks will have a tremendous capability to move data.
MEC has several advantages that core networks do not. The main ones are concurrency of components, lack of a global clock, and independent failure of components. In some cases, it may have a common goal, such as a massive multiplayer online 3D gaming and streaming HD media. Or, in the case of wireless communications between users, the goal is to make the path of the data between users as fast as possible, even with extremely dense traffic. MEC is ideal for that.
Because most of the data stay within the network, users experience much better responsiveness for all real-time remote monitoring or streaming use cases. Such networks also advance real-time remote broadcasting, a collection of virtual realities (XR), machinery control, vehicle-to-everything (V2X), and similar applications making communications as lag-free as possible.
The 5G element of MEC also enhances security (although it does not get quite the same attention as latency or other features). MEC enhances security just by the nature of the topology. While the technical explanation is somewhat lengthy, in essence, when devices enter a defined area with the MEC enhanced service it essentially allows the service to run on a virtual private network (VPN) without the need for any VPN setup. It is transparent to the mobile user and traffic to the service never exists on the public internet. However, that does not mean that this condition is an excuse to get lax on security.
Of all the MEC apps the IoX is likely to be the most challenging use case. IoX devices will have virtually unlimited scenarios. Some might consume a lot of bandwidth. Others will have minimal periodic data transmissions. And there will be a plethora of platforms from low-data rate to mission-critical medical and life-safety to autonomous vehicles.
In most cases, the challenge for IoX MEC will be to aggregate and analyze the data before it becomes actionable. And count on this being big data for many applications. For this AI/MI/ML (artificial intelligence/machine intelligence/machine learning) will be the underlying technologies and systems MEC will use to accomplish this. All of this in a concentrated area will allow for unprecedented functionality and capabilities.
One of the more visible use cases for MEC are sports venues. For some time now, stadiums, particularly football, have been test beds for accelerated and aggregated data and real-time streaming. In case you missed it, at this year’s Super Bowl, one of the Verizon commercials promoted having multiple, real-time camera views of the game via their 5G stadium platform. And there was a lot of 5G smack talk from both Verizon and T-Mo about 5G.
An early example of that was the 2019 Madrid Open tennis matches. MEC was used to offer fans, both inside and outside the stadium access to dozens of cameras, offering a VX-type of experience. Viewers were able to flip, seamlessly, between cameras and add elements such as digital scoreboards, TV, and social feeds as they watched the matches.
This has actually been going on for a few years now. The nuclear networking of sports venues to be able to offer over-the-top content within the confines of the venue. Until 5G came on the scene there was a Brundlefly of Wi-Fi, BT, and other technologies along with licensed spectrum to improve the fan experience. Now, 5G takes that to a new height.
Since then, this vector has picked up a great deal of traction, but only recently has 5G entered the scene. Frankly, this use case is a pretty easy challenge to overcome. The area is contained, all the users are on the same page and the environment has no real restrictions for hardware deployment (private grounds). However, other, more challenging scenarios, particularly where diverse technologies, topologies, and platforms exist, will experience higher levels of difficulty against increased complexities.
For example, for smart city deployments MEC will have to be augmented by things like massive machine-type-communications (mMTC), high-bandwidth media, AI, and its derivatives, big data and more to enable cities to aggregate and analyze data from millions of connected devices close to, and within, the confines of the deployment. This will require this advanced AI, enabled by 5G, to manage the diverse ecosystem of street lighting, parking and traffic control, personal safety, air and water quality, first responders, weather, and more.
Edge computing will be a game-changer, especially when married to 5G. It is the key component to enable and deliver a wide range of news and existing end-user services challenged by high bandwidth demands and latency.
But it is still evolving. For example, IDC predicts that by 2023 more than 50 percent of new enterprise IT infrastructures will be at the edge vs. traditional corporate datacenters. Today that number is around 10 percent. However, this again is not one of the more challenging scenarios.
We have a good grasp on what edge computing can do. However, we have a bit of a stretch ahead before we get all the details and bugs worked out and deploy multitudes of edge networks across all of the ecosystems that can benefit from it.
Verizon is joining forces with Microsoft to create new ways for enterprises to accelerate the delivery of fast and secure 5G applications, benefiting from reliable and low latency connections. Verizon’s on-site 5G Edge network integrated with Azure edge services can enable ultra-low latency, many times faster than the blink of an eye, which can help businesses tap into real-time data analysis and delivery.
Applications incorporating computer vision, augmented, mixed and virtual reality, digital twins or machine learning can be enhanced with 5G and MEC on the customer premise, helping transform the way industries such as retail, transportation, and logistics operate: Think of automated high-precision asset localization, tracking and positioning in manufacturing. In healthcare, the increased speed, reduced latency and high bandwidth connectivity of 5G networks could enable real-time precision medicine leveraging mixed reality and AI capabilities as well as seamless and fast sharing of large files to improve patient care.
The collaboration brings Azure cloud and edge capabilities together with Verizon’s on-site 5G Edge, a mobile edge computing platform designed to enable developers to build applications for mobile end-users and wireless edge devices with ultra-low latency. By providing on-site private 5G, businesses will realize increased power efficiencies and reduced costs of end user devices, while helping to address their privacy and security needs.
“We have built a network that provides real-world, 5G-enabled solutions TODAY,” said Rima Qureshi, EVP and Chief Strategy Officer at Verizon. “By bringing together Verizon’s 5G network and on-site 5G Edge platform with Microsoft’s expertise in cloud services, we will enable the development of the next generation technologies everyone has been envisioning.”
Logistics and supply chain solutions company Ice Mobility is already testing on Verizon’s on-site 5G Edge platform, integrated with Microsoft Azure. The company is using 5G and MEC to help with computer vision assisted product packing. By gathering data in near real-time on product packing errors, the company has the potential to improve on-site quality assurance and save 15% to 30% in processing time.
“We are especially excited to join Verizon and Microsoft to test how 5G and MEC can improve the quality assurance process,” said Mike Mohr, CEO of Ice Mobility. “They truly have listened to our needs to provide automated real-time quality oversight and feedback, which will enable us to cost-effectively launch unique new products, while maintaining the highest execution standards, significantly increasing throughput and reducing costs. And, this is just the beginning.”
“By leveraging Verizon’s 5G network integrated with Microsoft’s cloud and edge capabilities, developers and businesses can benefit from fast, secure and reliable connections to deliver seamless digital experiences from massive industrial IoT workloads to precision medicine,” said Yousef Khalidi, corporate vice president Azure for Operators at Microsoft.
Moving forward, Verizon will explore opportunities to co-innovate with Microsoft to deliver new value to industries ranging from manufacturing to healthcare.
EdgePresence has embarked on an aggressive effort to deploy low-cost edge data centers nationally at cell tower sites in tier 2 and tier 3 markets.
Through partnerships with Schneider Electric and its distributor Accu-Tech, EdgePresence is deploying fully integrated Schneider Electric EcoStruxure Data Center solutions, which are redundant 12-cabinet microdata centers that serve 20 to 30 key clients in these markets that need more connectivity and more options for redundancy.
Doug Recker, CEO of EdgePresence, whose background is in building, owning and operating data centers, has developed a two-phase approach that targets enterprises initially and then serves as a carrier-neutral solution to wireless carriers’ 5G edge compute needs.
“We have a business model that serves enterprises, and it can survive without 5G. We serve doctors’ offices, cloud providers, local hospitals and government,” Recker said. “I need revenue to sustain my business. I have employees to pay and investors to pay back. Enterprises are a very sustainable business model. When 5G comes, our data centers will already be in place and ready for the content providers and the carriers.”
When it comes time for the wireless carriers to build out 5G to tier 2 and tier 3 markets, Recker’s solution means carriers won’t have to build their own edge data centers. It makes it easier and quicker to deploy 5G.
“Carriers can buy access from me to the carrier hotel or for access to the internet at a fraction of the cost of building their own facilities,” Recker said. “No reason to deploy new shelters or upgrade their current data centers when they can save the capex by collocating with me. Just rent the space you need.”
Why Cell Tower Sites?
EdgePresence has signed a one-year contract with American Tower and has helped them deploy six data centers. It has plans to deploy at sites belonging to SBA Communications, Vertical Bridge and Crown Castle International.
“We need the ability to go anywhere,” Recker said. “I need space in all kinds of markets, and cell tower sites are my number one choice. Cell towers have the power and usually have fiber and a secure compound. It is easier to get permitted to deploy there. It is a whole new revenue stream that is waiting for them.”
These edge data centers also represent opportunities for tower services companies to deploy more equipment at the cell tower site and provide maintenance services.
Each structure has the same 12-cabinet redundant data center form factor, and the maintenance contracts will be the same, together with the same monitoring capability. Recker is looking to partner with one or more tower services companies to maintain the data centers.
“Maintenance is extremely important when we are deploying these data centers in third-tier markets around the country,” Recker said. We need people in place to who will fix them and maintain them. We went with Schneider Electric because it has a very low failure rate. The shelter must be cost-effective so we can make our money back quickly.”
EdgePresence has recently deployed a data center at a cell tower site in Statesboro, Georgia, and it built dark fiber back to the closest carrier hotel. Now, enterprises in Statesboro have access to more than 26 internet service providers, instead of one. They can use the facility for redundancy and collocation.
The key to edge compute is timing. If EdgePresence waits until 5G reaches the tier 2 and tier 3 markets to deploy its data centers, it might be too late. Another company may have already deployed a data center and be in position to fill that need. The company will deploy its next data center in Jacksonville, Florida, next week and then in Gainesville, Florida, the week after that.
“Our goal is to have 16 data centers deployed by the end of the year and have them 60 percent occupied,” Recker said. “Then it’s time to go turbo, and we will deploy more than 50 data centers in the next 16 months. If I am the first data center in a market, there is no need for another one. That’s why quick deployment is important.”
American Tower has teamed with a data center collocation provider, Flexential, to launch the Flexential Local Edge, a remote data center providing collocation services.
The first Flexential Local Edge remote data center will use American Tower’s edge data center at a tower in Atlanta, connected with Colo Atl, an American Tower subsidiary company that provides carrier-neutral collocation and interconnection services in a global telecom hub of Atlanta. Flexential plans to use American Tower edge data center infrastructure in Denver and Boulder, Colorado.
“This new arrangement is effectively a channel agreement, whereby Flexential will sell edge capacity to its clients at American Tower micro data center sites, which typically have nine racks and 96 kilowatts of total power,” wrote Eric Luebchow, a senior analyst at Wells Fargo Securities, in an equity research note. “It will be interesting to see how this symbiotic relationship between towers and edge data centers evolves in the coming years.”
Flexential’s remote data center allows customers to deploy hardware and software close to the end-user or application — at the edge, as it is called — in third-party managed edge data centers for the provision of low-latency communications, high security and comprehensive network services.
“Local Edge will push to use 5G and AI in a more meaningful way for traffic controls, autonomous vehicles, infrastructure monitoring and more,” said Flexential CEO Chris Downie. “We see this an opportunity in many markets around the country to provide a true differentiator.”
Other Tower Companies Already in the Edge Game
In 2019, SBA Communications purchased New Continuum Data Centers, which has a multitenant flagship data center in western Chicago and a fiber loop. Two years ago, Crown Castle International expanded its partnership with edge computing pioneer Vapor IO, making use of Crown Castle’s tower real estate and metro fiber.
Lately, with increasing adoption of work-at-home initiatives and video applications, telemedicine and remote learning as a means of limiting the spread of COVID-19 infections, more businesses seek to deploy low-latency applications to improve the overall customer and employee experience, which can be achieved through mobile edge computing.
“It’s still far too early to quantify the potential upside to these relatively minor investments, but they do signify the tower companies are watching the space closely and may be prepared to pounce if the opportunity presents itself,” said Luebchow.
Verizon Rapidly Rolling out Mobile Edge Computing
Last month, Verizon, teaming with Amazon Web Services, added three more 5G mobile edge computing cities — Atlanta, New York and Washington, D.C. In August, the mobile edge computing platform was rolled out by the companies in Boston and in the San Francisco Bay Area, and more cities will be launched by the end of 2020.
“Mobile edge computing moves the data and processing done by the applications and services we use closer to the end user at the edge of the network,” a Verizon spokesman said. “This shortens the roundtrip data needs to travel, reducing lag time, or latency. By moving AWS compute and storage services to the edge of Verizon’s 5G Ultra Wideband millimeter-wave network, innovators can develop applications with ultra-low latencies that will support next generation use cases ranging from self-driving cars to autonomous industrial equipment. Customers are already testing their edge solutions in AWS Wavelength service zones at the edge of Verizon’s 5G network in multiple locations.”
5G Americas has published ‘5G at the Edge,’ a white paper that explores new groundbreaking possibilities emerging from the combination of Edge Computing with 5G technologies. Edge Computing refers to locating applications, general-purpose compute, storage, and associated switching and control functions that are required to run them – relatively close to end users and/or IoT endpoints.
Chris Pearson, president, 5G Americas said, “Edge Computing locates processing power closer to where data collection actually takes place – nearer to the radio access network than the core, but it’s not one size fits all. As operators deploy edge computing, they will need to consider their architecture to address specific services, applications and use cases.”
This 5G Americas’ white paper explores Edge Computing’s role in the evolution of 5G architecture, the application of Cloud-native principles such as software defined networking (SDN) and network function virtualization (NFV), and identifies various methodologies currently being adopted for 5G applications. It covers detailed emerging use cases and outlines the stringent requirements needed to facilitate advanced mobility, compute, storage capabilities for emerging 5G wireless networks.
Additionally, this paper supplies an in-depth view of the various industry and open source initiatives defining emerging EDGE architectures. Overall, it defines the next generation Edge reference architecture and explores future directions in networking.
According to Pearson, “As 5G networks become more distributed, they will get more complex and need data processing that takes advantage of cloud-native principles like containerization and micro-services. Edge computing and network transformation will lead operators to enable new low-latency scenarios in Augmented Reality, V2X transportation, manufacturing, health, education and beyond.”
“A new reference architecture based on data centric technologies like analytics, networking and storage for edge computing-enabled 5G systems is being shaped that will have broad implications for how wireless networks operate in the future,” explained Rao Yallapragada, Director of Advanced Technologies for Intel and a co-leader of the paper’s working group.
“5G At The Edge” covers some pertinent topics:
According to Rao, “5G and Edge Computing are mutually reinforcing. Edge architecture’s need for low latency drives demand for 5G, and 5G’s growing availability increases the pull of workloads from the core to the Edge.”