Trials & Tribulations

Although the E-911 Phase II deadline is just 18 months away, standards committees still are grappling with several unresolved issues.

Having a cell phone makes it much more likely that you can call for help. But will the help be able to find you?

Today, if you can't identify your location, the best information available is the cell site that your phone is calling from. But that approach often identifies an area of several square miles. Under the FCC's E-911 Phase II mandate, by 2001, cellular and PCS providers must provide public-safety answering points (PSAPs) with much more accurate information about the caller's location. Although that deadline is less than a year away, there are still many questions about the most effective approach and whether the best that today's technology can offer will be good enough.

Network-based location systems rely on basic geometrical principles: Monitoring equipment at multiple locations analyzes the distance that the mobile's signal has traveled or its direction, calculated by determining the time that a signal took to travel from the phone to the monitoring point, or by using both types of information.

But in the RF world, nothing is simple. Signals fragment as they bounce off buildings, so they wind up arriving from different directions and with different delays. These phenomena create errors, which can be exacerbated by the alignment of the mobile relative to the monitoring stations. Even with more monitoring stations, there still will be significant errors.

Handset-assisted systems rely on the mobile using a built-in GPS receiver to help determine its location. Some systems, which truly can be called handset-based, are completely independent from the phone and can both calculate the position and transmit it directly to the PSAP via, for example, voice synthesis. However, this approach requires a complete GPS receiver plus additional circuitry attached to the phone.

Another solution is for a built-in GPS receiver to calculate and then transmit the position to the PSAP via the existing signaling network. This approach reduces the cost by eliminating some of the hardware and software requirements.

A third approach is for the wireless phone to perform minimal analysis of the GPS signal and send the GPS pseudo-ranges — the estimated distances between the GPS satellites and the phone — to the network for analysis. This approach minimizes the impact on the phone but requires more sophisticated network equipment. A new TIA standard, IS-801, describes a protocol for a CDMA phone to communicate this information to a base station for network analysis, and similar work is under way for both TDMA and analog phones.

Needles in Haystacks
The Phase II Report & Order initially appeared to eliminate handset-based solutions because of the requirement to provide the location of all mobiles and not just those with location capabilities. But in late 1998, the FCC indicated that it would consider waivers. The Third Report & Order sets explicit rules for handset solutions. Although the original rules required that 67% of mobile locations be reported within 125 meters of the actual location, the new rules distinguish between network and handset solutions.

Network solutions now must locate 67% of mobiles within 100 meters and 95% within 300 meters, while handset solutions must locate 67% within 50 meters and 95% within 150 meters. To address the concern that handset solutions would leave many people unlocatable, the FCC required that possibly as early as April 2002, all new handsets should be location-capable, while all handsets would have to be location capable perhaps as soon as the end of 2004. The uncertainty inthese dates is because the clock starts only when a PSAP is ready to receive the new types of location data.

Standards to the Rescue
With only a handful of field trials to go by, it's tough to judge whether a network or handset solution is best. Network solutions will face multipath problems in dense-urban areas and in rural areas will have a hard time getting the necessary multiple fixes on a signal. On the other hand, GPS-based solutions rely on a good view of multiple satellites, which are difficult to see in dense-urban areas, inside buildings and under other shading, such as canopies of trees.

Aside from a pure handset solution with location information sent via a synthesized voice, network modifications are necessary. Regardless of how the caller's location is determined, that information still has to get from the wireless network to the PSAP. TIA standards subcommittee TR-45.2 and ATIS committee T1P1 have been struggling with solutions to this problem, being documented in the joint TR-45.2/T1P1 document PN-3890.

One solution for conveying the location information is to enhance the SS7 ISUP call-setup signaling to transmit the location as part of call setup. Although this approach is conceptually elegant, many systems still rely on older MF tone-based signaling that can't be modified to transmit additional information. It also requires that 911 call setup be delayed until the location is determined. If some locations take many seconds to produce, the delays in call setup may become unacceptable. After all, PSAPs are anxious to get callers' locations, but they are far more anxious to talk to them.

Another alternative is to leave call setup unmodified and transmit location via a non-call-associated network, such as one based on X.25 or IP. This approach allows more time to deliver the information because it will take several seconds for the PSAP to determine the emergency type before dispatching, which requires location information. This solution also allows location updates, helpful when the initial location proves to be highly inaccurate.

Another network consideration is whether the location should simply be delivered to the PSAP or whether it should be used to control routing. In many places, cell sites cover multiple jurisdictions, each of which may have separate call takers. Precious time can be lost when one town's PSAP has to forward the call to another town's PSAP. The probability of this occurring could be reduced if the caller's precise location is mapped against jurisdictional boundaries, with routing of the call and delivery of the location being directed to the PSAP.

Besides providing multiple network-interconnection types, PN-3890 also has to be able to process data generated from different types of network and handset positioning systems. Some wireless-service providers might first implement a network-based solution and then migrate to a handset-based solution. The network will have to recognize which mobiles can use which solution. Different subsets of the PN-3890 protocol will be required in each case.

Phase II is a major technical challenge exacerbated by a lack of cost-recovery mechanisms, wireline service providers' delays in upgrading their equipment and many PSAPs' inability to fund equipment upgrades. As playwright Noel Coward wrote, "Some day I'll find you, moonlight behind you," but for most 911 callers, that day still is many moons in the future.

Crowe (crowed@cnp-wireless.com) is a wireless-standards consultant and editor of Cellular Networking Perspectives, a wireless-standards and -technology bulletin.