Strength in Numbers

Despite the ads touting nationwide coverage, no wireless carrier has seamless coverage in every corner of the United States. The Wireless Consumers Alliance, which sparred with the industry over ways to improve 911 access, estimated that in urban areas, as much as 10% of any analog service area is uncovered. In suburban areas, that figure is 25%; for rural areas, it's 33%.

But combine both analog carriers' coverage areas in any market, and the holes diminish significantly. That was the logic driving the debate over how to improve the chances of completing 911 calls made from wireless phones. In May, the FCC required all analog handsets manufactured after February 2000 to be able to route a 911 call to the other carrier if it can't connect using the primary carrier. The goal is to ensure that callers will reach help even in areas where their carriers have poor or no coverage.

The FCC provided three solution options -- automatic A/B roaming - intelligent retry (A/B-IR), strongest signal and selective retry -- and left it up to each carrier and vendor to implement their choices. All three are primarily handset upgrades and transparent to the network because most, if not all, networks complete 911 calls without first checking roaming agreements, blacklists and authentication. So it's possible that in February, a carrier could have all three solutions operating in its markets simultaneously.

"I don't really care what method the handset vendors implement," said one carrier executive who had just begun discussions with his vendors. "What we have to ensure is that one of the three is in every handset that we buy."

PROS & CONSThat ambivalence toward solution choices is a far cry from the fractious debate just a few months earlier. The solution that drew the most criticism was strongest signal, where the handset would use the carrier with the strongest control-channel signal. Many in the industry denounced strongest signal for the potential cost of its intellectual-property rights. But perhaps the biggest gripe was technical: The solution chooses the strongest control channel, which doesn't necessarily mean that a voice channel robust enough to support a call also is available.

Standards also will determine which solutions are chosen.

"The adequate/strongest-signal solution is the most difficult to implement," said Phil Hester, Ericsson director of product marketing. That's because the other two solutions are better supported in the current standards.

"Adequate/strongest signal will require a little bit of standards work," said Jim McGarrah, BellSouth Cellular director of network services. "Because standards is a consensus process, typically that's not an overnight process. On the other hand, a lot of the people who have been involved in this discussion through the Wireless E-911 Implementation Ad-Hoc Committee (WEIAD) also are involved in the standards committees. So I suspect thought has already been given to how standards might be modified to accommodate these."

The strongest industry support appears to be for A/B-IR, where the handset first tries to connect through the preferred carrier, such as the subscriber's carrier or its roaming partner, before trying the other carrier. If the alternate carrier doesn't have a usable signal or vacant voice channel, the handset tries the preferred carrier again and then alternates between the two until the call is completed or the subscriber gives up. Once the call completes, A/B-IR monitors the voice channel. If the call is dropped, the handset would automatically try again.

Selective retry is a sort of manual A/B roaming: The caller uses a button to toggle between the two carriers until the call goes through. Besides a software upgrade, selective retry also requires a hardware change to add the button.

Getting compliant handsets to market by February 2000 was a concern during the debate, when the industry proposed a 12-to-18-month window to implement the solutions. The FCC acknowledged that a tighter schedule might disrupt vendors' product cycles but argued that the public-safety issues were compelling enough to warrant a 9-month roll-out. Handset vendors participated in WEIAD, and most have had one or more solutions under development well before the ruling.

"We feel like we've already created a solution that deals with the issue," said Arnold Gum, Qualcomm senior product manager. "It's something the industry and Qualcomm have been working on for a while. I don't think that there's going to be a problem with meeting the deadline."

Any winnowing likely will come after vendors and carriers have had time to study each solution's real-world performance.

"I seriously doubt that all of our vendors will end up doing it the same way," said one carrier executive. "So what will happen is a couple of the methods -- not necessarily all three -- will get out there. Then as we see what happens, people will either modify the method, or one will be disqualified because it just doesn't seem to fit."

Cost and public perception also will determine which solutions survive.

"What the consumer demands will get implemented," said Trinh Vu, Siemens senior product manager. "For a high-end handset, I would expect automatic retry would be the feature. Then on the low-end handset, probably (selective retry). It's different market segments."

Cost also apparently isn't a concern.

"It's pretty minimal," Gum said. "I think we see that as part of our job in providing a handset. I don't really foresee any increase in price passed along to the carrier."

One unresolved issue is callback. Suppose that a 911 caller connects via the alternate carrier because the preferred carrier doesn't have coverage in the area. If the caller is in an E-911 Phase I service area, the carrier that completes the call is responsible for supplying emergency personnel with the caller's phone number and the cell site where the call originated. But unless the preferred and alternate carriers have a roaming agreement -- and that's rarely the case, for competitive reasons -- emergency personnel probably wouldn't be able to call the 911 caller back. Even so, few in the industry and public-safety community see this limitation as enough to scuttle the entire concept, especially when callback on wireline 911 is subject to its own vagaries.

"Although this might be seen as a shortcoming to the FCC proposal, I believe the frequency with which this might occur would be miniscule," said Eric Sorensen, SCC Communications product-marketing manager. "We also have to remember that there are a lot of other reasons that (emergency personnel) may not be able to reach a wireless subscriber via their (number). For instance, they may have turned off their phone or traveled out of a covered area."

Industry estimates show that carriers completed 98,000 911 calls each day in 1998. How many more didn't complete is anyone's guess, but all three proposals should help.

"None of these by themselves are an ideal, 100% guarantee of a call completion," said BellSouth's McGarrah. "Of course, that's due in large part to the fact that RF is such an imprecise science, influenced by terrain, foliage and atmospheric conditions. All of these are attempting to provide alternatives to just the conventional set-up-a-call-only-with-my-preferred-carrier approach to maximize the possibility of call completion. But none of these are slam dunks."