SAN FRANCISCO – AT&T is bolstering its work with edge deployments but notes that it will take a lot of collaboration – and a few years – for those 5G edge benefits to hit the market.
That sentiment was expressed by Igal Elbaz, senior vice president for wireless network architecture and design at AT&T, at the carrier’s Spark event this week in San Francisco. Elbaz explained that a robust edge deployment will be necessary for the carrier to deploy immersive services that will differentiate the carrier’s 5G network from its current 4G LTE deployment.
“Latency is important when you think about immersion,” Elbaz said, who noted that current network architectures cannot provide the low level of network latency required to support more robust 5G use cases. “5G is all about a differentiated service. It needs to be more than just LTE.”
AT&T’s network architecture is structured with three levels: the network layer, which is the somewhat traditional wireless network; the SDN layer that is required to control more complicated services; and the edge.
“This is the distributed cloud,” Elbaz explained. “This is what brings those experiences closer to the user and really differentiates 5G.”
He noted that while the premise of edge deployments is understood, there is still work needed in terms of actual implementations. “It’s going to take some time,” Elbaz said. “The concept is well understood and the need is understood.” He added that the next step was generating enough collaboration across the ecosystem to make it happen.
“It’s a collaborative model with the ecosystem so we can help shape it,” he said. “There are a lot of questions, and we don’t want to be in a position where we are solving these ourselves.”
Elbaz reference the recently launched Akraino Edge Stack Project as “the blueprint that can foster innovation” with the development community. AT&T provided the seed code for that project that is currently housed within the Linux Foundation.
He also expressed the need for “creative engineering collaboration.” This includes the sharing of some of the vast amount of network data and insight from the carrier’s network.
“We are definitely open to sharing some of that intelligence of our network in a smart way,” he said. “We can provide guidance that no one else can.” This includes insight into network latency and throughput in real time. This could be a significant step as operators have historically been averse to sharing that deep customer insight. “All things need to be on the table,” Elbaz said.
Ebalz also said the carrier would eventually tap into all of its available spectrum assets to support 5G services. This includes low-band spectrum that is currently supporting its 3G and 4G networks. He explained that it was likely the carrier could follow the model it has used with LTE in which it has re-farmed spectrum used for its 2G and some of its 3G services to support 4G.
“It’s not just the high frequencies,” he explained. “We need to take advantage of all of our assets.”
AT&T yesterday said it would tap into the unlicensed 3.5 GHz Citizens Broadband Radio Service (CBRS) spectrum band to support a fixed wireless 5G service set to launch later this year. The carrier’s current mobile 5G plans involve use of millimeter wave (mmWave) spectrum in the 28 GHz and 39 GHz bands.
The CBRS spectrum will provide better coverage compared with the mmWave spectrum bands. The carrier will use CBRS to support rural broadband deployments to expand broadband services to the one-third of the country that still lacks a high-speed internet connection.
“It’s where we see an immediate need and a way to serve that need,” Elbaz said. “It really should be fundamental for all Americans to have access to broadband.”
Igal Elbaz, senior vice president for wireless network architecture and design at AT&T, speaking at the carrier’s Spark event this week in San Francisco.