Enhanced 911 puts GPS into cell phones

SKYAID
New
Mission
Overview
Details
Medical
Watch
Heart attack
Stroke
World health
Emergency
Cost effective
Media
- Site Map

SKYCAR
Details
Overview
VTOL
Airline
Military
Transportation
Images
- Site Map

Search

Translate
8 languages

Enhanced 911 service spurs integration of GPS into cell phones
By Junko Yoshida, EE Times Aug 15, 1999 (9:28 PM) URL: http://www.eetimes.com/story/OEG19990815S0005

SANTA CLARA, Calif. — With a ruling due next month from the Federal Communications Commission on Enhanced 911 (E911) technologies, the race is on to tuck global positioning satellite (GPS) capabilities inside cell phone handsets. Software and hardware technology vendors believe an FCC decision to expand the definition of E911 to include handset-based technologies could be the catalyst that propels GPS into high-volume markets, enabling the rollout of E911 services "one handset at a time," said Ellen Kirk, vice president of marketing and strategic planning at GPS-software house SnapTrack Inc. The alternative is to overhaul entire networks to handle E911, using a scheme for pinpointing by triangulation among cellular basestations the location from which emergency cellular phone calls are made.

Semiconductor designers, meanwhile, hope to pack GPS cores into the RF and baseband chips in cell phones. With an eye on this imminent market, SiRF Technology Inc. (Santa Clara, Calif.), a fabless chip company specializing in GPS technology, will announce today a new architecture designed for more-accurate positioning, faster signal acquisition and better performance when applied in tough locations like deep foliage and urban canyons. SiRF engineers are to detail the architecture, which will be offered as both a chip set and intellectual-property core, at this week's Hot Chips Conference at Stanford University.

Competitor SnapTrack (San Jose, Calif.), meanwhile, claims momentum is building to fold its unique, server-based technology into cell phones. Under a licensing deal recently announced with Texas Instruments Inc., the partners will jointly integrate SnapTrack's network-assisted GPS technology with TI's DSP platforms for digital wireless systems.

Proponents claim the handset-based approach to E911 can be deployed with much greater accuracy than network-centric schemes, well before the Oct. 1, 2001 deadline set by the FCC. They claim the FCC's original rules, adopted years ago, were built on the assumption that wireless carriers would upgrade their entire networks for E911, primarily with technologies such as radio triangulation.

"One of the big advantages of handset solutions is that we can begin providing [E911 service] coverage much sooner," said Kirk of SnapTrack. Alternative schemes, she said, would require the conversion of an entire network, entailing expenditures of time and money and placing a big burden on wireless carriers.

Into this domain comes the SiRFstarII architecture. It consists of three components: a GPS signal processor integrated with a full-function 50-MHz ARM7 CPU, plus 1 Mbit of extended-data-out DRAM on-chip and enhanced GPS capabilities; an RF chip incorporating an intermediate-frequency filter; and software to provide a flexible system architecture for standalone GPS-based products.

"The new GPS silicon solution has been developed from day one to be available both as a highly integrated chip set for standalone consumer applications as well as an intellectual-property core to be integrated into platforms such as cell phones," said Kanwar Chadha, founder and vice president of marketing at SiRF.

SiRF will sell the combination of the digital chip, fabbed by Samsung, and the RF chip, manufactured in NEC's 0.5-micron BiCMOS process, at $29.95 in volumes of 10,000.

While SiRF offers a fully featured, autonomous GPS solution in the form of chip set and core, SnapTrack offers carriers and handset vendors software that uses the network to aid in GPS tracking. The company claims that its server-assisted scheme improves upon conventional GPS performance by sharing processing and database functions between the mobile GPS receiver/processor (the client) and a remote infrastructure (the server and reference network).

The world's largest wireless carrier, NTT Docomo in Japan, will integrate both the SnapTrack and SiRF technologies into its Docomo Location Platform, upon which Docomo plans to build a variety of location-based services for its cell-phone users. The platform, scheduled for launch in December, aims to enable "new classes of pedestrian navigation services" using the wireless infrastructure married with GPS. The success of GPS on the cell-phone market depends very much on new services those wireless carriers may come up with in the future.

No handset takers

However, no cellular handset companies have yet announced plans to take up either technology. SnapTrak's Kirk expects this situation to change as soon as the FCC's new ruling is handed down in September. SiRF's Chadha, meanwhile, pointed out that Nokia is one of his company's key investors.

As SnapTrack and SiRF pitch their technologies to OEMs and carriers, several leading GPS technology vendors, such as Motorola Inc. and Trimble Navigation Ltd. (Sunnyvale, Calif.), are virtually invisible in this segment of the GPS/cell-phone convergence market. "I'm very surprised that traditional GPS vendors — Motorola, Trimble and [others] — have so far all failed to address the needs of what may become a majority of the GPS market in the United States over the next few years," said Andy Fuertes, senior analyst at Allied Business Intelligence Inc. (Oyster Bay, N.Y.).

Fuertes projected that the GPS market will expand from 7.6 million units in 1999 to 162 million units in 2004. "The biggest force behind the market growth is the E911 mandate, while GPS is also expected to grow in PDA, automobile and PC applications," he said.

Technology issues aside, the uncertainty over who is to shoulder the cost of E911 appears to be causing a stalemate between wireless carriers and the Public Safety Answering Point (PSAP), which represents the call centers that take 911 emergency calls. PSAP needs to upgrade its system to translate latitude and longitude information into digital maps, but the agency and carriers — faced with the prospect of wholesale network conversions of their own — are bickering over their expected cost-recovery scenarios.

"The chances of Public Safety Answering Point meeting the current FCC mandate by the Oct. 1, 2001, deadline are very slim [to] none," said Kirk of SnapTrack. "It physically isn't going to happen."

Kirk believes a network solution is possible. But, she said, "By getting the handset solutions added to the mix of choices for wireless carriers, carriers can enjoy competitive pricing and performance offered by competing technologies."

The handset camp has come together in the Advanced E911 Coalition, an industry group composed of companies like AT&T, Qualcomm, Sprint PCS, AirTouch and SnapTrack. The coalition is proposing an option to the FCC in case a handset solution — GPS-based or otherwise — is chosen: that carriers must, within 24 months of the FCC's adopting an order, ensure that 50 percent of all new phone activations are capable of handling so-called "Phase II" location services; within 36 months the number rises to 95 percent. Phase II, as defined by the FCC, means that carriers should provide the location of all 911 calls at an accuracy of within 125 meters.

SiRF's Chadha believes GPS can do other solutions one better in this regard. A case in point is the company's new architecture. In a move far exceeding the FCC mandate of an accuracy of 125 meters, SiRFstar II can home in on a caller's location within 2 to 15 meters by taking advantage of two upcoming GPS infrastructure enhancements.

The first is the Wide Area Augmentation System (WAAS), a recent initiative of the Federal Aviation Administration to improve the accuracy and integrity of GPS signals. Such new signals are being tested now and are scheduled to begin broadcasting next year. The second is the Department of Transportation's national network of beacon-based, differential stations, called National Differential GPS (NDGPS). Also due for launch next year, NDGPS was designed to extend the U.S. Coast Guard's differential network.

SiRF added two separate blocks on its newly designed digital chip, with the idea that each could run a different software algorithm to accommodate WAAS and NDGPS independently.

Those capabilities could put the SiRF approach a step closer to SnapTrack's software solution in terms of location finding. When asked to compare its technology with conventional GPS, SnapTrack has always pointed to its ability to operate inside buildings and in urban canyons to determine location within a few seconds, and to share circuitry with cell-phone handsets. The latest improvements in SiRF's technology could neutralize that advantage.

Signal snapshot

Many of SnapTrack's stated advantages simply come from how the technology is set up. When SnapTrack is activated, the wireless network sends an estimate of the location of the handset to a server. The server tells the handset which GPS satellites are nearby. The handset takes a "snapshot" of the GPS signal, calculates its distance from all satellites in view and sends this information back to the server.

The server software performs complex error correction and calculates the caller's precise latitude, longitude and altitude. In a 911 call, the server sends the information to the PSAP. For other location-based applications, the server can send the coordinates to a third-party service provider, a dispatcher or back to the handset. SnapTrack claims the process takes just a few seconds, whereas conventional GPS receivers can take several minutes.

SiRF has responded to the need to improve GPS accuracy in obscure locations by adding multipath-mitigation hardware to its new digital chip. "We've added hardware in a very low-level DSP block, so that we can separate out direct signals and differential reflected signals to improve accuracy," said Greg Turetzky, product-line manager at SiRF. Because SiRF's solution still needs to receive both direct and reflected signals, it is more suited for urban-canyon environments than for indoor use.

SiRF's new silicon also features an improved satellite search engine that the company says increases the speed of the initial satellite signal acquisition by a factor of eight. "This allows a 10-second decrease in cold start," or cases when the GPS handset, using an inexpensive crystal, has no idea where it is, Turetzkey said.

As for shared circuitry between GPS functions and cell phones, SiRF claims very little difference between what's required by SiRFstarII and by SnapTrack. SiRFstarII is expressly designed to go inside a cell phone, essentially meeting the phone vendors' strict demand to add GPS capabilities "with no additional cost, size, power, CPU and memory," Chadha said.

The embedded EDO DRAM inside SiRF's GPS signal processor is designed as on-chip memory dedicated to GPS navigation. The digital chip also packs a satellite signal tracking engine, which eliminates the need for a high-rate GPS interrupt. That means that 90 percent of the throughput of the chip's ARM CPU remains available for non-GPS tasks.

To get GPS into a cell phone, both SnapTrack and SiRF require a radio-frequency chip and antenna. SnapTrack also requires a DSP and separate memory. Although it may share the DSP already inside a cell phone, extra glue logic is likely to be needed so that its GPS function can access separate memory directly.

In a manner similar to SnapTrack's network-driven scheme, SiRF has added a network-assisted positioning solution to its bag of tricks. SiRFLoc helps the chip take advantage of the precise orbit data and acquisition aids provided to wireless network carriers whenever the GPS is in a weak-signal environment. Carriers are now developing standards for wireless assistance for position location.