Blog Post

2014: The year AT&T’s mobile network goes small

Stay on Top of Enterprise Technology Trends

Get updates impacting your industry from our GigaOm Research Community
Join the Community!

For the last year, AT&T(s t) has been poking and prodding small cells in its labs and out in the wild. The tiny little base stations will allow AT&T to surgically insert capacity into its HSPA and LTE networks in high-traffic areas like malls, stadiums and public areas where its customers are using their smartphones the most.

Those cells will bring a lot of benefits for consumers. They’ll offer up to greater speeds to customers in dense areas such as downtown financial districts and other high-traffic zones. They’ll bring stronger signals to the dead zones between cells on AT&T’s network. And they’ll bring greater coverage and capacity in difficult-to-penetrate indoor locations. Finally they’ll free up room on the macro network, giving customers in between small cells a better experience. But first AT&T has to work out the kinks.


The man in charge of the rollout, AT&T Associate VP of Small Cells Gordon Mansfield, has been testing out small cells from numerous vendors in every configuration imaginable. He’s mounted them to walls, on ceilings and hung them off utility poles. He’s connected to them microwave transmitters, cable coax and even U-verse fiber-to-the-node modems.

He’s put one vendor’s cells uncomfortably close to a competitor’s cells just to make sure they play nice. He’s placed the tiny transmitters at the edges of its networks where signals are the weakest, but he’s also dropped them right smack in the middle of a macrocell to see how they behave.

Gordon Mansfield AT&T

It’s a tricky business, Mansfield told me in a recent interview, because AT&T can’t just plop down a small cell anywhere. Since AT&T is using the same frequencies for small cells it’s using for its umbrella macro-network, interference becomes an issue. If the signal clash between the big and small networks is too great, customers won’t get faster connections. Instead, their phones would only see white noise.

But after a year of trial and error, Mansfield thinks AT&T has got it down to a science, quite literally. AT&T Labs scientists have created a software tool called HetNet Analysis and Resource Planning (HARP), which largely automates the process of determining where best to place cells while mitigating potential interference in the network.

With the help of HARP, AT&T has deployed HSPA small cell networks in 18 states. Those deployments range from corporate high-rises to big indoor gathering spaces to outdoor pedestrian zones. And those various scenarios have given AT&T the data and confidence it needs to begin a widespread rollout of small cells in both HSPA and LTE flavors, Mansfield said.

AT&T's HARP Tool at work
AT&T’s HARP Tool at work

“Right now I can say we’re in a fast walk,” Mansfield said. “But we’re right at the cusp of stepping on the gas and going into a full-blown sprint.”

That means were going to see small cells popping up in a lot of places in 2014, though Mansfield wouldn’t reveal a specific number. AT&T, however, has announced plans to add 40,000 of the little buggers to its network by the end of 2015 as part of its Project VIP IP-transformation plan.

enterprise small cells
Ericsson’s Radio Dot

A lot of eyes are on AT&T and Mansfield, who happens to being pulling double duty as the global chairman for the Small Cell Forum. AT&T’s VIP initiative will be one of the largest rollouts of public-access small cells in the world. Those small cells will be a key building block in AT&T’s future heterogeneous network, or HetNet, which will tie together cells of all sizes and Wi-Fi into one big interwoven system. Such networks will not only have to seamlessly move subscribers between many more nodes, they could even connect devices to multiple nodes simultaneously.

With the addition of new self-optimizing network (SON) technology from Cisco-Intucell, AT&T’s network may even cease being a static grid and become a moving organism. Cells will anticipate where users are going, dynamically growing and shrinking as those users move through the network.

Designing today’s small cell networks might sound difficult, but planning tomorrow’s networks will be akin to completing a jigsaw puzzle in which the pieces’ shapes are constantly changing.

2 Responses to “2014: The year AT&T’s mobile network goes small”

  1. The fact that AT&T is deploying small cells in the same frequency spectrum as the macrocells will result in interference between the macro and small cell layers that will limit the benefits of small cells, especially outdoor ones. They will have to upgrade to LTE-Advanced to solve this problem. We anticipate that small cells will pick up on volume no earlier than 2015.

  2. It would be smart of att to overhaul their wifi network with small cells. Interference could be an issue depending on location, but a multi mode small cell (LTE Hspa+ and wifi) small cell in like a Barnes and Noble makes sense especially considering the backhaul is there for the most part. I wonder what AT&T’s plans are for there HSPA+ network? There’s a lot more potential and headroom there than say an old cdma like Verizon or Sprints but they haven’t said anything about it. Any thoughts on any of that?