LNP Juggling Act

Mobility always has been wireless' main advantage over wire-line, but it became a mixed blessing in July 1996, when the FCC ruled that wireline and wireless carriers would have to allow subscribers to keep their phone numbers when they changed providers.

Wireline LNP debuted a year ago in the 100 largest MSAs. Under LNP Phase I, wireless carriers must be able to complete wireless calls made to ported wireline numbers, which currently total about 400,000. Under Phase II, by March 2000, wireless carriers also must be able to port their own numbers. In that environment, a wireless subscriber could churn to another wireless carrier or even a wireline carrier but keep the same number.

The FCC gave wireless carriers more time to prepare for Phase II, and for good reason. One area where mobility complicates LNP is by forcing the separation of the MIN and the mobile directory number (MDN).

"When a roamer enters your market, you don't know what the phone number of the subscriber is any more," said Lou Grilli, GTE TSI manager of business development.

In an LNP environment, if that roamer calls for help, the carrier could wind up giving the wrong information to the public-safety answering point (PSAP).

"The MIN, which is typically what's given to the PSAPs, is not the callback number," Grilli said.

That change poses a serious hurdle for an industry that prides itself as a safety asset. "Heaven forbid they dial 911," said Chris Kostenbader, Bell Atlantic Mobile (BAM) director of network switch and facilities planning. "Under the FCC order where we have to give the 20-digit information, they're going to get the wrong callback number."

Wireless LNP already is under way in Great Britain, and in Canada, PCS carrier Microcell Telecommunications is pushing regulators to mandate LNP based on a survey showing that 40% of wireless subscribers were likely to switch carriers if they could keep their numbers. In the United States, however, the CTIA and most wireless carriers want to limit the industry's LNP obligations to Phase I until the deadline for broadband PCS build-out expires in 2004. By March 16, 1999, the FCC must rule on the CTIA's petition to postpone Phase II. When asked what carriers should do in 1999 to meet the Phase II deadline, Kostenbader joked, "Get a gun and go shoot themselves might be one of the things."

SNOOZE & LOSE?How much LNP will cost is difficult to pin down because much of the hardware and software, particularly for Phase II, is still coming to market. One barometer is where and how much wireline carriers are spending on LNP.

"What we're hearing from the wireline side is that the back office is going to cost the most," said Frank Salm, AG Communications Systems product marketing manager for INgage solutions. "I'm hearing incumbents saying it's costing somewhere between $500 million and $1 billion to implement LNP, and yet the LNP solutions themselves are just in the single-million-digit range. So the bulk of that effort has to do with all those back-office systems. I expect we're going to see similar situations with the wireless carriers."

Carriers can ignore Phase I by simply allowing the LECs to route the calls, but that approach can be costly because the LEC will charge the carrier for each database lookup, referred to as a "dip," necessary to determine whether the number is ported. Still, defaulting to the LEC could be a viable short-term option for carriers serving markets with few ported numbers.

"That's really a market-by-market decision," Grilli said.

That decision also hinges on how much the LECs will charge for completing calls. Grilli said tariff filings have shown dip charges ranging from $.0005 to $.006. As a result, when planning for both Phase I and II, carriers need a solid forecast of the number of calls to ported numbers that they expect their subscribers to make. With that estimate, they can determine, for example, how many transactions per second their SS7 networks must be able to accommodate. At the end of 1998, only 400,000 wireline numbers were ported, but that figure is misleading.

"You'd think, 'How many calls are going to be made to those numbers?' That's a very low transaction rate," said Dave Cox, director of product marketing for Tekelec's network-switching division. "The first assumption you'd make is that one pair of STPs or SCPs will handle all the traffic I need. That's not true. Any time any number in an NPA-NXX is ported, then that entire (10,000-number) NPA-NXX must be marked as ported. Therefore, every call to that NPA-NXX must do a dip to the LNP database. That increases the number of transactions per second greatly."

Another issue is the percentage of wireless calls made to wireline numbers. If the current 80% estimate is correct, defaulting to the LEC could be a costly Phase I solution.

"If 80% of your subscribers' calls terminate to wireline, somebody's got to do an LNP dip," Cox said. "The LEC's going to charge them to do an LNP lookup even when the call is not going to a ported number. What's happening is that wireless carriers using that scenario are getting some pretty large bills every month for dip charges for their basic traffic. So they need to consider implementing LNP right now. It's going to cost them X amount of dollars in capital expenditure to implement LNP, but they're going to be doing their own dips. Therefore, they won't be paying on every call."

A DREAM DEFERRED?LNP couldn't come at a worse time for carriers, which will spend 1999 and early 2000 exterminating the Y2K bug, building out their networks and grappling with other costly mandates such as E-911. Judging by what several carriers say privately, Y2K has pushed LNP to the back burner.

"Most of the vendors, because of Y2K, are looking to have this functionally available in the mid-1999 time frame because they want to have it in and proven before Y2K," said AG's Salm. "What we're going to see toward the end of 1999 and possibly the first couple months of 2000 (is that) most wireless carriers are going to stop deploying major new infrastructure because they want to focus on the Y2K issue."

If there's any consolation, it's that many LNP upgrades such as increased SS7-network capacity clearly are investments in the future. One example is how Phase II forces the separation of the MIN and MDN. The unified MIN/MDN likely would have to be scrapped anyway for global roaming.

"We recognized that as part of the international standard we're going to migrate away from that," said BAM's Kostenbader. "There's going to be the new International Mobile Station Identification. So in the future, there's going to be a departure from using the same number for dialing and locating the phone. All we did when we came up with the LNP solution is recognize that that's a good solution. We just advanced the use of that idea and went to the MIN/MDN separation."

The MIN/MDN separation might not be as much of a chore as it first appears.

"Most carriers support phones without MDNs such as data phones for UPS, cellemetry and Aeris," said David Crowe, Cellular Networking Perspectives editor. "So they may be farther ahead than they realize."

Another issue is whether LNP will help the industry realize its dream of replacing wireline. Some carriers scoff at the notion that LNP will make people more inclined to switch wireless carriers or abandon wireline.

"If you go to the man on the street today, people are going to look at you with a funny look and not know what (LNP) means," said a spokesperson for one major carrier. "Where is the need for this outside the fact that the FCC wants to mandate this?"

A TO-DO LIST FOR 1999Carriers that plan to meet the March 2000 Phase II deadline will have a busy 1999. The process begins with determining long-term goals.

"Back up and look at the big picture," said Tekelec's Cox. Carriers should "look at planned steps that migrate them to the network they want to have rather than making isolated decisions around things such as LNP or fraud or whatever."

A solid LNP Phase II implementation plan includes:

* Teamwork. LNP isn't just an engineering issue. It affects every department and the entire business process. Create an implementation team that includes representatives from all departments.

"I'd highly recommend that you have representatives from your third-party suppliers," said Dave Breda, Pittiglio Rabin Todd & McGrath (PRTM) consultant. "They should be very aware of the deadlines."

* Planning. Create an implementation plan that outlines LNP deployment, an integration schedule and the effect on resources and other projects such as Y2K and E-911. Rather than one deadline, "you should have a half-dozen interim milestones between now and the end of the year," Breda said. "It distributes the risk along the way."

* Cost assessment.

"The cross-functional implementation team is a good way to assess the cost because you have subject-matter experts in each of the areas," Breda said. Accuracy is key because if traffic exceeds forecasts, costs will increase and quality of service (QoS) could suffer. One example is the SS7 network, which will see increased traffic and message sizes.

* Testing. LNP adds more elements to the network, so there are more things that potentially could go wrong.

"It's quite possible that their call processing is going to work, but it's the back office that's going to cause them problems," said AG's Salm. Testing early and often will catch problems before they snowball.

One possible test is having employees port numbers between you and another carrier over a period of several weeks. If major reliability problems crop up, you want to have enough time to assess those reliability issues, get together with your implementation team and write that petition to get it in and filed to the FCC in plenty of time so that they can rule on it.

"The FCC has been pretty clear that they want to see proof that there's an impact to the QoS in meeting these deadlines," Breda said. "If you can show that through testing, then you're more likely to get your petition granted. I'd say by at least November or December 1999, you need to be well into testing and not just from a systems' standpoint but from end to end."

* Education. The public isn't clamoring to port numbers, but sooner or later, they'll encounter LNP. That means the implementation budget should include resources for educating both the public and employees. Suppose a subscriber is having problems with his phone.

"Somebody calls in, and they ask, 'What's your phone number?'" said BAM's Kostenbader. "They may have an old phone that doesn't have the new register in it. They'll mention it as (908) 256-9811. Our repair technician might think they mean that's the directory number (MDN) when in fact they were giving the MIN. You've got that confusion factor that's going to permeate even voice interaction among people."

IS THERE A BUSINESS PLAN?LNP's costs will be tough to swallow for an entrepreneurial industry that's still building much of its infrastructure. One comparison is E-911. Although the E-911 Phase II deadline is October 2001, vendors spent much of 1998 unveiling enhanced services that would take advantage of location technology. So far, LNP hasn't spawned a similar market. Yet some remain cautiously optimistic about LNP's potential as a platform for enhanced services.

One possibility is prepaid.

"There's a huge, untapped source of revenue opportunity out there," said AG's Salm. "That's an example of an application that we develop that can run on the same SCP as the LNP solution. LNP is probably a more attractive way for adding services into the network (than E-911) because it introduces the IN infrastructure, the full SS7 (network) and the network elements to be able to handle all the services."

Finding a business plan for LNP involves looking at the difference between an infrastructure that will support LNP and one that also will support enhanced services.

"I think that it is a dangerous fallacy to believe that an LNP database will support other services," said Cellular Networking Perspectives' Crowe. "This is going to be a high-performance database that won't be able to meet its performance goals if it is bogged down with additional bells and whistlesthat aren't really necessary."

If new revenue streams from enhanced services won't defray LNP costs, what will? In the end, they'll probably be passed on to average subscribers.

"It's not fair, but I don't see any other way around it," Crowe said. "You can't charge customers leaving your system because it would be viewed as anti-competitive, and you can't collect anyway. And for competitive reasons, you probably can't charge customers porting into your system."

Passing LNP's costs on to subscribers puts more pressure on carriers in an increasingly competitive marketplace -- and jeopardizes wireless' dream of replacing wireline.

"They say that every cloud has a silver lining," Crowe said. "They also say that the exception proves the rule. Well, this is the cloud without a silver lining."

One wild card is calling party pays (CPP), which some think would complement LNP.

"The only barrier that was taken down (by LNP) is that you don't have to give up your number," said Kostenbader of BAM, one of the most vocal supporters of forbearance. "But that hasn't been a really big barrier to date with churn rates the way they are. CPP may change that. With CPP, I'd be more apt to give out my phone number, so I'd be more interested in keeping it if I changed carriers. CPP will help facilitate portability. It will help lower that barrier."

Some carriers might look to savvy marketing as a way to make the best of what's shaping up to be a bad situation. One possibility: highlighting the ability to port from carrier to carrier or even wireline to wireless.

"Why do you have to deal with this as a problem?" said GTE TSI's Grilli. "Why not deal with this as an opportunity? If we're going to allow people to leave our network and keep their numbers, doesn't that also mean we can allow people to come into our network and keep their numbers? Churn is bad if they're churning out. Churn is good if they're churning in."