A lot of attention has been focused on Sprint's network and its new CEO's determination to bring in customers with new pricing plans of late, but behind the scenes another major transformation is taking place within Sprint that is getting less attention, even though it ultimately affects every part of the big carrier's business and its future success. (See Sprint CEO: Price Cuts First, Best Network Next .)
As part of its NetworkVision 2.0 strategy, Sprint Corp. (NYSE: S) is currently moving to an automated process of tracking its network inventory, working with OSS vendor Netcracker Technology Corp. to define and put in place a new way to identify and track all the moving parts of its end-to-end networks. As part of that effort, Sprint is determining which legacy processes and systems it can leave behind, even as it actively builds out not one but two new wireless networks -- its enhanced 4G LTE network and the Spark network. (See Sprint Sparks to Reduce Churn, Save Unlimited , Sprint Sparks Up Vendors for Faster 4G LTE and Sprint Chooses NetCracker for OSS Network Vision.)
"At the same time you are doing all those things, you have to make sure there is minimal or no impact on the customer," says John Dye, director of IT application development at Sprint. Members of Dye's teams are participating with network engineering teams of each domain and members of the IT staff to incorporate new OSS automation processes in its new networks from the outset rather than later, as an afterthought. It is a major endeavor for both Sprint and its vendor.
NetCracker has been involved in many network transformation processes but the Sprint effort is a bit unusual, says Ed Feingold, the vendor's director of strategy.
"They are focused on the network inventory piece, which, in most cases, wouldn't be all that exciting," he says. "But Sprint is different because, first, they are working at extremely large scale and, second, they are doing three things at once: creating a new centralized network inventory repository that over time will decommission some of their legacy systems, but also building new systems to support the 4G network vision and now Spark."
Why inventory management?
Knowing what is in the network doesn't seem to be that big a deal -- unless you have acquired companies, run multiple lines of business, and migrated traffic from one network to another -- all of which Sprint has done. As Dye describes it, moving to a single view of the Sprint networks, which are constantly and automatically being updated and reconciled, is not only important to meeting customer demand and providing a quality of service customers expect, but also to protecting both the capex and opex budgets.
"We need to have a strong view of what is really in the network and not just what was designed or planned for the network," he says. "The currency factor, the ability to discover the network and be able to reconcile all that data and keep that current for everyone is critical."
As demand for data in particular skyrockets, Sprint must target its infrastructure investments to where they are needed and know that existing investments aren't being stranded or under-utilized. At the same time, the carrier needs to accurately track the "before and after" of new network and equipment deployment.
Sprint initially used NetCracker to help in the OSS transition to the all-IP world, then applied the vendor's inventory model to its radio access network. Now as it builds out the 4G LTE and Spark networks, it is using the same model to ensure consistency in building and deployment activities, so that as things are added, the inventory is kept current and all elements can be validated and accepted.
"With that, you can go straight through to a service assurance model, and that drives the ability to do accurate performance management," Dye says. "Performance management drives many things out there -- how you look at capacity management, how you look at performance and the availability of the network; also, how you look at driving customer experience management to make sure what you have done and what you are doing going forward is the right thing for customers."
By tracking all this information and making it available across the network organization as Sprint goes into a new deployment, the carrier targets its network spending more carefully, on both the capex and opex sides, Dye says.
"Obviously we want to keep our capex down as we are spending all this deployment money so that we don't over-provision because someone doesn't have the right capacity and configuration data," he comments. "And we are also keeping opex down because we can use our partners effectively -- minimize the amount of effort on their parts, which keeps their costs down, saving money which they pass back through to us."
Another place where Sprint expects to save money is in retiring legacy OSSs that are no longer needed. But determining what can be retired and managing that is an exacting process in itself. It often involves extracting legacy data from existing systems and populating that in the new system, NetCracker's central repository, in a different way.
"We are constantly looking at that portfolio, assessing it as we look forward to the Spark network and the future technologies there, doing application rationalization, and determining how we can consolidate that best," Dye says. "We have a long history of legacy apps, and we definitely want to get those out and reduce our opex for maintaining those."
That is often a staged process, he adds, with detailed transitional plans to phase out applications partially or completely as they aren't needed. By eliminating systems that aren't needed, Sprint can also reduce licensing costs and overall operations budgets, Feingold notes.
The hard part
NetCracker right now has a team of people in place working alongside Sprint to implement the new inventory management process. One of hardest parts of this entire process is change management -- training people to do things differently on newer systems.
New systems mean new user interfaces, new data entered differently, and new functions for back office staff, and adoption often goes slower than expected, Dye says. "Getting people used to the system is always one of the biggest challenges -- it's a people challenge, not a technology challenge."
Ultimately, it is likely that fewer people will be needed but those people will need a broader range of skills, Dye says, encompassing both telecom and IT training. Sprint's new CEO has also said job cuts are inevitable and will based on necessity.
Sprint will look to incorporate virtualization in its OSSs where and when it makes sense, Dye adds. He sees some value in reducing the number of boxes deployed through the software-centric model but admits concerns about customer impact. "It's a little too early to know what impact this might have," he comments. "I want to see this stuff working in the lab before we use it."
— Carol Wilson, Editor-at-Large, Light Reading