DALLAS -- Network Virtualization & SDN Americas -- AT&T is concentrating on the middle of the edge to build the network intelligence needed for emerging 5G apps, Josh Goodell, AT&T VP of Edge Solutions, said Wednesday.

AT&T is focused on building virtualization and intelligence into the customer edge and network edge for 5G applications, Goodell said, kicking off the Network Virtualization & SDN Americas conference here.

"This is not science fiction for us. We are in the process of delivering it today," Goodell said.

Putting intelligence in the customer edge means moving compute to the customer location, which allows for local processing of data on site, providing privacy for companies that need it, and application level controls, Goodell said. Manufacturing is one application demanding customer premises intelligence.

Intelligence at the customer edge is integral to a project where AT&T is working with Rush Medical, a connected hospital, to enable real time image processing and analysis by medical staff at the location. Additionally, facial recognition monitors patient pain thresholds to alert doctors and nurses to change their therapeutic approach to the patient. And the hospital is using AR and VR for doctor training. "This is happening today. We are deploying this solution as a real example, as we speak right now," Goodell said.

Figure 1: AT&T's Josh Goodell.

Deep dive into real-world issues and virtualization deployment challenges with industry leaders. Join Light Reading at the annual Network Virtualization & SDN Americas event in Dallas through Thursday. Register now for this exclusive opportunity to learn from and network with industry experts. Communications service providers get in free!

Light Reading wrote about AT&T's collaboration with Rush Medical previously.

The other part of the network middle is putting compute in the network itself. "We call this network edge compute," Goodell said. AT&T is pushing compute deeper into metros, in central offices, to optimize wireless and wireline networks, enabling 5G low latency.

"The big difference between this approach and on-premises is that this will span an entire metro, whereas the other approach is centralized to a location," Goodell said. Network edge compute can bring roughly 20 millisecond latency to a 200-300 mile radius.

As an example of the type of application that can be permitted by network edge compute, Goodell cited AT&T's partnership with Vorpal, a company developing technology to monitor rogue drones to track their location, and also triangulate back to find the operator. Goodell said that seemed to him at first like a niche application, but drones are an emerging security threat, piloted by saboteurs and terrorists. Goodell noted that drones may have played a role in the recent attacks on Saudi Arabian oil facilities.

AT&T started its transformation to virtual networking by laying out a vision in 2015, Goodell said. And we've followed AT&T's virtualization transformation closely over the years. Most recently, last week we reported that AT&T is on track for 100% core network traffic virtualization next year, with 75% of MPLS tunnel traffic on the core network now controlled by SDN. AT&T is leaning on open source to implement virtualization -- it says it has no choice.

Today, AT&T is moving to a virtualized network in over 70 countries. The next step is to make the network "truly intelligent," Goodell said.

5G is the big change driver today, Goodell said. 5G provides great benefit over existing technology: 100 times the throughput of LTE, providing multiple gigabits per second. 5G also provides for "ultra responsiveness" -- in other words, low latency.

That network performance permits revolutionary applications, Goodell said.

"In the not too distant future it will be possible for a surgeon sitting in Boston to operate on a patient in Boise, Idaho, 2,000 miles away in an operating room using 5G technology," Goodell said.

Massive device connectivity is another new capability of 5G transformation, Goodell said. While WiFi can support hundreds of devices per access point, and LTE can support thousands, 5G can support millions. That may sound like overkill, but emerging applications need that kind of connectivity. A single manufacturing location might support 35 million sensors in one building.

Related posts:

— Mitch Wagner Executive Editor, Light Reading