AT&T taps big data for network management

AT&T (NYSE:T) is using big data and network analytics to better understand its customers' network experience and thereby improve on that.

On a typical day, AT&T says it can measure 1.9 billion quality checkpoints from its wireless network. In a typical hour, it's taking about 30 billion network quality data measurements across its wired and wireless networks.

Through AT&T's program, called Service Quality Management, AT&T Labs has invented "breakthrough technologies that use sophisticated analytics to make sense of this massive volume of network data and discern what customers are experiencing," according to a blog post.

"These technologies have transformed how we manage our network--driving smarter decisions and resolving issues quicker than ever before," the post says.

One primary technology the company developed is what it calls TONA, the Tower Outage and Network Analyzer, which uses real-time data and historical measurements to capture the ripple effect of a network disruption. It guides in deciding which disrupted cell towers to fix first, shortening the duration of events for the greatest number of people. It factors in population and nearby towers that can handle the problem and uses real-time data and historical measurements to capture the effect of a network disruption. AT&T says TONA has created a 59 percent improvement in identifying customer impact.

Other tools in AT&T's arsenal include Mercury, which uses analytics to help understand the health of the network after planned maintenance, and MINT, which helps pinpoint the exact location in the network that is generating an issue and why.

While it's easy to see how all operators can benefit by using big data to improve network quality, AT&T's use of big data may be especially pertinent given the Project Velocity IP (VIP) program it announced back in 2012. AT&T committed to deploying more than 40,000 low-power small cells by the end of 2015 as part of that program, along with new macrocells and distributed antenna systems (DAS) throughout its network.

The network troubleshooting tools also are notable considering how AT&T was not prepared for the demands that the iPhone so famously brought to its network in the early days of Apple's entre into wireless.  

Gartner estimates that big data analytics can impact a communications service provider's financials by $300 million a year. It identifies four areas as ripe for big data analytics: network optimization, customer engagement, operations and applications and new business models.

This AT&T video summarizes its usage of big data analytics in wireless network maintenance. (Source: AT&T / YouTube)

For more:
- see this AT&T blog
- see this white paper

Related articles
AT&T cheers FCC's forthcoming moves to speed up wireless infrastructure deployment
AT&T's De la Vega: We've already taken 'hundreds of thousands' of iPhone 6, 6 Plus pre-orders
Marian Croak on why and how AT&T is moving to SDN, NFV and an all-IP future with Domain 2.0
AT&T completes Disney DAS and small cell rollouts
AT&T to expand LTE network to 300M POPs, deploy at least 40,000 small cells