AT&T wises up about outages thanks to TowerScan

TowerScan, which began as a summer intern project during the summer of 2011, has already become a key tool used by AT&T (NYSE:T) to assess the impact of tower outages on the actual customer experience.

The technology was deployed in a little over a year by AT&T's Global Network Operations Center (GNOC) and enables the company to quantify the customer experience across a complex outage, according to a recent article on the AT&T Labs Research website.

TowerScan is much more efficient than using manual methods that can be fooled by network redundancy, which normally ensures mobile customers are within reach of at least two cell towers that can handle their calls. "While you would expect customers on a failed tower to be impacted, they may not be impacted if they get moved to another tower while customers on nearby, still-functioning towers may be impacted when their tower takes on traffic from the failed one," said the article.

This type of issue cannot be captured by manual methods of assessing outages based on the number of failed towers or population density, according to AT&T.

AT&T: A cell tower failure may (or may not) affect customers--both those customers on the failed tower and those on nearby towers.

For that reason, TowerScan uses network data to analyze not only the failed tower but at all towers within a 10-mile radius to see which customers are being impacted and how. A TowerScan analysis is initiated either manually or automatically if a scan of outage alerts returns a positive result.

To tell the difference between increased traffic loads that are caused by a nearby outage rather than the normal start of rush-hour traffic, TowerScan applies statistical time-series analysis to compare a tower's current number of customers with the expected number for both the time of day and day of the week.

The new technology has given AT&T insight into the actual impact of outages, helping the operator prioritize repairs. Interestingly, outages with a few failed towers may have much larger customer impacts than outages with a significant number of failed towers.

"The impact depends on how the outage occurs--if failed towers are less clustered, for example, the outage may have less impact than an equivalent outage of failed towers tightly coupled together, demonstrating that network metrics and population density provide a poor approximation of customer impact," said the company.

The issue of tower outages, network resiliency and prioritization of repairs was highlighted in the aftermath of Hurricane Sandy. At the height of the storm, 25 percent of cell sites in 158 counties in 10 states from Virginia to Massachusetts were inoperable. It took about a week for most service to be restored by all four Tier 1 carriers in the affected areas.

AT&T said it will soon incorporate TowerScan's information about network outages into its customer-service workflow. This should help service reps evaluate whether a customer-reported problem is likely caused by a tower outage or by an unrelated problem and, thus, more accurately predict when service will be restored.

As its researchers continue to enhance TowerScan, AT&T said it hopes to develop an understanding of the impact outages can have on customers across the 2G, 3G and LTE technology layers.

Customers may be moved from layer to layer during an outage but the cascading effect is not well understood, said the company. For example, it is unclear whether moving an LTE customer to a remote tower that still has LTE is better or worse than moving the customer to a closer 3G tower.

For more:
- see this AT&T Research Labs article

Related articles:
FCC holds hearings on Hurricane Sandy's damage to wireless networks, seeks answers
Hurricane Sandy exposed flaw in public-safety LTE plan
Hurricane Sandy: Wireless carriers restore most disrupted service
T-Mobile CTO: Hurricane Sandy delayed some 1900 MHz HSPA+ launches
AT&T, T-Mobile open networks for Hurricane Sandy victims in NY, NJ

Article updated on March 8, 2013, to reflect a change made by AT&T in its original article.