Can You Get There From Here?

Ask a vendor or carrier about least-cost routing (LCR), and more than likely you'll hear, "Huh?" The confusion is as much about the how as it is about the why. Many PBX systems have the ability to pick the cheapest route on a call-by-call basis by choosing a long-distance carrier based on the time of day or day of the week. But adapting that model to the wireless world is another matter.

"The ability exists in today's network," said Brad Cooper, AirTouch network-systems engineer. "But I don't think there are a lot of carriers taking advantage of the time-of-day aspect, where your more expensive trunks during peak hours are utilized more during off-peak times to capture any discounts from the LEC while satisfying any contractual minimums."

One concern is whether more choices actually will translate into more savings. All but the smallest carriers have enough traffic -- and thus clout -- to negotiate decent rates with IXCs and other carriers.

"It's not something that you'd want to do or even have the ability to do because it would cost more money to do it on a day-by-day basis," said Keith Paglusch, Sprint PCS senior vice president of operations. "You have to satisfy yourself that whatever agreements you're entering into, you're getting that" same level of savings.

Any savings might be offset by the cost of administrative tasks, such as compiling and continually updating a database of rates.

"If there were some kind of network-management system that used data warehousing or something that could extract data, and if you had multiple connections to different carriers, there may be a business case for it," said Marybeth Selby, Telcordia (formerly Bellcore) Wireless Technology Solution executive director. "But I just don't see it."

Determining the cheapest route in the second or two it takes to complete a call is another hurdle. Even if there were a national clearinghouse, roaming, taxes and other fees, all of which vary significantly state by state, are among the hidden costs that could take the L out of LCR.

"It would be a nightmare for them to interface to tables and then send commands to their switches," said Samir Marwaha, Inet OSS marketing-strategy manager. "I'm sure that it's possible, but it comes down to a business case: Would they be able to justify spending a huge amount of money, and then what would they gain from that?"

FROM POINT A TO POINT B -- EVENTUALLY If LCR isn't viable, streamlining routing processes is. As sophisticated as SS7 is, it's also somewhat of a Rube Goldberg machine, at least in its current incarnation.

One example is what's known as "tromboning" or "hairpinning." Imagine a salesman from Kansas City roaming in Calgary, where his 3 p.m. appointment calls his wireless phone to cancel. The Calgary system would route the call to his Kansas City carrier, which would route the call back to its Calgary roaming partner. The result is a call involving two long-distance legs.

One solution is a roaming agreement that allows the Calgary system to query the Kansas City HLR, get the location of the Calgary roaming partner's MSC and route the call directly to it. That approach uses a single local leg instead of two long-distance legs. That's technically possible today, except for one thing: the North American Numbering Plan.

"How does the Calgary MSC recognize that the number I just dialed is a Kansas City mobile and not a Kansas City landline?" said David Crowe, Cellular Networking Perspectives editor. "This would be virtually impossible and, with LNP, absolutely impossible. Consequently, the Calgary MSC treats the number like landline and routes to Kansas City. By the time it gets there, it's too late to do anything."

Two things could salvage it. One is Release to Pivot, a relatively new SS7 feature originally designed to accommodate LNP. With it, the Kansas City MSC would communicate with the Calgary system to get a temporary local-directory number (TLDN) for routing.

"This TLDN then would be sent to the originating switch to cause the call to be routed to the switch where the mobile is roaming," Crowe said. "That would convert two long-distance legs into a single local leg."

The other option, already under consideration for calling party pays, is to set aside blocks of numbers for wireless.

"You would know from the numbering plan that you were calling a wireless number," said Randy Snyder, Synacom Technology executive director of marketing. "That's very helpful (because) the switches know that it's a mobile number and has to be treated differently."

Implementing that scheme uniformly across North America is one barrier, but perhaps the biggest is the politics that come with eliminating long-distance legs in favor of local legs.

"Every once in a while, somebody brings this up in standards (meetings)," Snyder said. "It's been going on for years. But guess who shoots this down? All the companies that make money off the long-distance legs of those calls."

That likely will change, thanks to competitive pressures and the need to cut overhead in order to offer 1-rate plans. Carriers with large geographic footprints still could eliminate tromboning within their own networks. Those that own their long-distance legs or have close ties to long-distance carriers have an advantage because both provide greater control over pricing and resources.

"It's all owned by the same company, so it would be to their benefit to save on system resources," Snyder said. "If they're charging 10 cents a minute flat rate, why use long-distance legs?"

FOR WHOM THE BELLS TOLL As the number of wireless-to-wireless calls increases, direct trunks between wireless carriers serving the same geographic area also can streamline routing and cut costs by eliminating the public switched telephone network (PSTN) middlemen. Every call requires a database lookup to determine who owns the number. If it belongs to the partner carrier, the switch routes the call directly to it. One wild card is the number of wireless carriers in the area.

"Wherever traffic warrants, direct trunks can be useful," Crowe said. "However, as the number of competitors increases, the percentage of calls to any particular switch will drop."

Bypassing the PSTN also provides better performance because transcoding isn't necessary. But the biggest appeal is not having to pay wireline carriers to carry the call. As a result, direct trunks are starting to become common, but carriers are reluctant to acknowledge their existence.

"There is an interconnection type for wireless-to-wireless," said Snyder, who helped write IS-93. "It was defined by the standards because the carriers needed it. There are carriers that do direct trunking. It is very efficient, and they do save money."

Those savings don't necessarily have to be passed on to subscribers. LNP and the proliferation of new area codes obscure the relationship between a phone number and a geographic location.

"The area codes don't mean anything anymore because there are so many splits and overlays," Snyder said. "Who's going to know that 341 is only a mile away? So if they do direct trunking, they still can charge what they were charging before and save money by not having a LEC or IXC carry the call to the other cellular carrier."