Breakthrough Ideas (continued, with [additions] in square brackets)
GpsOne
Position location services are potentially valuable to users, manufacturers, and service providers. Market studies by Allied Business Intelligence and Cahners In-Stat Group estimated that wireless location service would generate $7B to $8B in new revenues for wireless operators over the next four years. The North American FCC E-911 Mandate sets additional performance standards that must be met by October 2002. Yet, with one exception to the rule, current methods of pinpointing locations of wireless users remain too slow and too inaccurate for either adequate response to an emergency or for requested services such as travel maps and traffic conditions or a list of nearby restaurants.
In January 2000, Qualcomm acquired its subsidiary SnapTrack because of its breakthrough Personal Location Technology, called Assisted GPS (A-GPS), that combines GPS satellite technology and wireless handset technology to produce consistently rapid and highly accurate fixes of positions. This proven end-to-end solution distributes the processing requirements between SnapTrack’s SnapSmart server software and a position location chip in mobile terminals by communicating over the wireless network. The SnapTrack system is agnostic to wireless air-interfaces, whether terrestrial or satellite, operating in any thin-device regardless of generation, band, or mode, on any type of wireless or pager network. Anytime and anywhere you can place a wireless call, you can rapidly pinpoint your location using their server-assisted-GPS system.
The distributed steps in the action-path are: (a) caller makes location request; (b) wireless network sends location of closest cell site; (c) SnapSmart server tells the handset which GPS satellites to locate specifically where; (d) handset reads GPS signals, calculates distance from each satellite, and sends information; (e) Smart Server’s sophisticated DSP algorithms introduce error corrections when calculating the longitude, latitude, and altitude of the handset; and, (f) in case of E-911 calls, the server contacts the most appropriate public safety access point, or, for service requests, contacts the third-party service provider or Internet portal.
Whereas traditional GPS technology has difficulty acquiring signals when signals are attenuated by 5-10 dB, the enhanced SnapTrack system can acquire and provide fixes even when there is 25 dB signal attenuation. This increased sensitivity enables SnapTrack to provide position location fixes indoors, in automobiles, in urban canyons, or under dense foliage, markedly increasing the yield of successful fixes to attempts. Whereas traditional GPS technology requires from 30 seconds to several minutes to provide a first fix, SnapTrack provides an initial fix in a few seconds in worst-case environments and in less than a second under an open sky. Because SnapTrack technology provides a snapshot of data, typically in only 0.1 to 1 second, for Location On Demand, and then turns off, it dissipates a small fraction of a handset’s power. SnapTrack is accurate within 5 to 20 meters.
This difference in performance between traditional GPS and Assisted GPS is a function of SnapTrack’s innovations that removed the four essential location functions from the GPS-terminal and distributed them across the DSP-receiver, wireless network, and server. For instance, Qualcomm’s MultimodeGlobalPositioning6200 coprocessor supports a position location solution for GSM/GPRS/UMTS that provides a superior solution in terms of cost, sensitivity, time-to-first-fix, and power consumption. (However, the flexibility of SnapTrack permits its software to run on any handset’s DSP processor.) By deeply integrating a location coprocessor into the phone’s components, the cost is minimal compared to requiring separate boards or discrete devices. Also, Moore’s law promises continual cost reductions with every MSM cycle. As volume scales, experience-curve-reductions in cost will also follow. Currently, about 3 [10?] million gpsOne terminal are in use, supporting nearly 100 location-based services.
The Smart Server provides key information and a navigational solution that simultaneously uses a terrain-elevation data base to correct altitude, mitigates multipath effects by optimizing information from sequential deflected signals, cross-correlates strong signals to correct pseudo ranges coming from weaker satellite signals, and corrects for atmospheric delays. This server-assisted GPS minimizes handset costs and improves acquisition times, sensitivity, and accuracy. Best of all, this system works where others fail: inside buildings, cars, and deep urban canyons. Other than adding the Smart Server location software, no infrastructure modifications are required for A-GPS in IS-95 or cdma2000 networks.
Qualcomm’s gpsOne position location solution integrates the SnapTrack server-technology with a position location coprocessor to form a hybrid A-GPS system that also uses Cell-ID. Qualcomm’s gpsOne is both the technology and market-share leader. Today gpsOne is deployed in 3 million handsets. More than 15 handset models with gpsOne are available in the U. S., Japan, or South Korea. The gpsOne solution is available on MSM3300, 5100, 5500, and becomes a standard feature on all 6xxx chipsets, except the entry-level 6000.
As a technology, gpsOne achieved the following firsts: 1st integrated wireless baseband with GPS capability; 1st integrated wireless RF with GPS capability; 1st GPS-based solution to work reliably indoors; and 1st integrated commercial end-to-end CDMA-based tracking service using QCT chipsets and the SnapTrack Location Server (by CoCoSecom, a Japanese security company). Not only is gpsOne the lowest-cost integrated wireless GPS solution in the world, but also it is the only commercially proven positioning solution for 3G wireless.
Only carriers in the U.S. who opted to use gpsOne or SnapTrack solutions comply with the FCC E-911 Phase II milestone requirements. In fact, according to Telephony’s Lynette Luna, “AT&T Wireless Services is pushing technology vendor Nokia to foot a $2.2 million bill the FCC slapped on the carrier in May for failing to sell E-911 capable GSM handsets on time. …The unusual action highlights brewing controversy about the commercial readiness of a location technology format known as enhanced observed time differences of arrival, or E-OTD.”
Qualcomm and SnapTrack seek to become a ubiquitous complete solution for position location. Seeking to extend its solution to Europe and Asia, QCT has a white paper devoted to an analysis of location technologies for GSM/GPRS/WCDMA networks. These networks bring additional problems to position location, and the white paper compares and contrasts alternatives, including Cell-ID, E-OTD (Enhanced Observed Time Difference) for GSM/GPRS networks, OTDOA (Observed Time Difference of Arrival) for UMTS networks, A-GPS, and hybrid A-GPS + Cell-ID, which is Qualcomm’s preferred solution.
It is not surprising that yield, speed, and accuracy are strong determiners of the value of location-based services. Back office wireless systems have always collected information about cell or sector locations, which is called Cell-ID. Cell-ID produces 100% yield, but it is variously inaccurate and inconsistent because it relies on information from the single base station closest to the mobile, regardless of where you are physically located within that cell, where the diameter of a GSM cell can be anywhere from 2km to 20km. Thus, Cell-ID information is accurate, say, between about 500m and 5km, answering the location question, “What town?”
E-OTD triangulates the position of the mobile by monitoring transmission bursts from multiple BTSs to measure time shifts between arrivals of GSM frames. Because precise time information is critical and GSM is an asynchronous system, this requires the addition of a Location Measurement Unit (LMU) infrastructure, at a ratio of 1 LMU to 1.5 BTS sites, throughout the network. To provide network-wide coverage, this could require 1000s of LMUs at or near BTSs, 1000s of installation approvals, and significant network planning to estimate, discover, and correct the RF impact. Either network expansion or cross-network roaming would require similar coordinated changes in location software and added LMUs in either extended- or roamed-to-networks.
Special software is required for legacy mobiles to support E-OTD, which requires that the subscriber take the phone to a specified location because the software cannot be downloaded over the air (no BREW). E-OTD also requires that many data messages be exchanged, with constant updating. Thus, this format for location information uses more network bandwidth than A-GPS. In rural areas, there is a poorer yield, and accuracy varies with BTS density and location. So, the two-dimensional solutions (no altitude) provide an average accuracy of 100m to 500m, answering the location question, “What part of town?” .
A variant of E-OTD, OTDOA operates only in WCDMA networks using a similar triangulation approach from three Node B stations. Two intersecting hyperbolas defined by observed time differences in arrivals estimate location within 100m to 500m. However, the low power requirements used with spread spectrum reduce its yield, the ratio of the number of successfully calculated positions to the number of attempts, because the low power requirement for WCDMA limits the number of sites the handset can see and use. Yet, OTDOA shares the same limitations with E-OTD: time dependency drives need for LMUs, poor yield in areas without 3 nodes, poor accuracy along linear networks, say, along highways, multipath degradation, compatibility only on UMTS network and not GSM/GPRS, uses up network bandwidth for messages and timing, and poor expansion and roaming that require location software and extension of location measurement units for timing synchronization. In fact, OTDOA may not be a viable solution because a study in Germany found that base station visibility for obtaining longitude and latitude was available only between 28% and 36% of the time.
As I speculated in Part II when discussing the asynchrony problem, according to Qualcomm (p. 13, my emphasis), “Since WCDMA networks require new infrastructure, many networks will be synchronized to optimize communications [reference to Qualcomm’s synchronization white paper] or designed so they can be synchronized at a later date by adding timing equipment to appropriate network elements.” This confirms my speculation that Qualcomm’s UMTS chipset will have built in, if not turned on, capacity for GPS synchrony, a Trojan Horse. It simply solves too many problems (and increases the probability of sales) for Qualcomm to omit the possibility of easy future synchronization when so much potential functionality is lost by the failure of UMTS to synchronize at a network level.
[The above paragraph appears to be a good example of the sort of speculative errors that, I, as a non-technologist, may make. In SI post # 32195, poster rkral kindly pointed out my error. Ron’s main point was that “cellphone chipsets can’t make cellsites send synchronous signals,” and “Qualcomm doesn’t build any W-CDMA cellsite chipsets.” He also did not care for the metaphor of “Trojan Horse” if it implied any “evidence of trickery or deceit by QCOM or Dr. Irwin Jacobs.” I hope that it did not because I did not intend any such implication.
At the recent Annual Meeting, in response to a question about synchronization in WCDMA, Dr. Jacobs’ response, as I interpret it, said that several changes in the design of WCDMA, nominally to improve functioning, but, in fact, degrading it had occurred, including the switch to asynchronous handovers. Quoting from Q1000’s valuable, but non-literal, transcript of the call, Dr. Jacobs continued,
“It is possible to do it [handoff] without the synchronization, but it does have some cost to it. So, in designing the WCDMA specification, one of the things we have done is to point out that there are advantages—battery life, better handoff capacity—in having synchronization. The infrastructure currently being designed does not include the synchronization software capabilities, I would say, in general. And, so that’s something that can be added in. The ability to make use of that in the handset is not just a software as opposed to a hardware issue [sic?]. An upgrade of the software will then at some time in the future allow that advantage to be achieved.”
I am not exactly sure of either what Dr. Jacobs said or intended here, particularly in the sentence marked with a “sic.” However, I believe his general comments on the need to offer complete solutions across bands of spectrum, generations and the like, what I call the “harmonization strategy,” which includes a process of continual upgraded improvements, means that Qualcomm is deeply committed to offering an eventual solution to the asynchronous problem in WCDMA, perhaps, by downloading synchronization software and adding timing equipment. I simply was premature in thinking that this had been achieved and was resident on Qualcomm MSMs and CSMs, but Qualcomm undoubtedly has the capabilities to solve this asynchronous problem. And, I would not be surprised if Qualcomm had in fact created the virtual equivalent of a WCDMA CSM as part of its effort to test the interoperability of its own WCDMA MSM5200, 6200, and 6250 chipsets.
I hope that technologists swill continue to correct any misconceptions that I hold or may spread. Better yet, they might try their hand at writing the descriptions of the various technologies that I try to describe whenever I have them hopelessly muddled.]
This failure to synchronize creates a second barrier to roaming across networks by adding OTDOA’s need for synchronization to permit position location to the barriers of dropped calls in asynchronous “soft” handoffs. For carriers that want to roam seamlessly across networks, only synchronization to the GPS standard facilitates network effects. Qualcomm offers this synchrony now, plus global harmonization is on the way.
Riding in wearing an easy-to-locate white hat, Qualcomm’s Assisted-GPS operates across GSM, GPRS, and UMTS networks, whether asynchronous or synchronous, by using three satellites to triangulate the latitude, longitude, and altitude of a position. Distributed assistance from a position location server improves accuracy, pinpointing positions within 5m to 50m, answering the question, “What street corner?”
Assistance improves signal sensitivity significantly, increasing yield above the other alternatives. Accuracy becomes consistent across geographies and networks. First-fix, minus network latency, is usually 5 to 10 seconds; additional fixes profit from first-fix information, becoming more rapid. The power drain is reduced. Roaming becomes excellent because it only requires similar position location server software in the roamed-to network. The system is efficient, using minimum network bandwidth and capacity. Expansion is easy. Compatibility is excellent because it supports all old or new networks. Snap-Track or gpsOne only costs a few dollars per phone. It doesn’t drain the battery or influence the form factor. In a real-world test of E-911 emergency services in Denver Colorado, location fixes were generated with nearly 100% yield and accuracy well under the required 50 meters in obstructed environments and within 3 meters in open-sky environments.
The use of hybrid A-GPS with Cell-ID offers all the advantages of A-GPS, plus improves yield still further.
The common characteristic shared by all five approaches is that accuracy, speed, and yield are affected by the degree of time precision. In 2G or 3G CDMA systems, precisely synchronized networks produce precise times. In asynchronous systems, the need for precise time requires adding LMU infrastructure. QCT provided an estimated cost after one year of switching to either an E-OTD system or an A-GPS system. The white paper concluded (p. 22, original emphasis), “In this example, the cost for the E-OTD-based system is about 2.5 times that of the A-GPS-based system.”
Therefore, Qualcomm maintains that for GMS/GPRS deployment A-GPS is the better choice. With E-OTD, there are large upfront costs for LMUs to provide timing, and the network will still need augmenting by A-GPS for advanced position location services. In contrast, A-GPS is less expensive and easier to deploy, and is compatible across all air interfaces. QCT (p. 28) concluded, “it’s clear that an A-GPS-based solution, and in particular the A-GPS hybrid, offers the best performance, provides the best ROI, operates equally well on 2G, 2.5G and 3G networks, and enables the most extensive list of location services.”
If you don’t want to be found, SnapTrack ensures your privacy through its Location on Demand opt-in feature that resides in the handset under your control. This is unlike E-OTD or OTODA network systems, which have the inherent ability to track a mobile device without the express consent or even awareness of the consumer.
In a 2000 study by the Strategis Group of European Wireless Location Services, the following four categories and listed examples were identified: (1) Trigger Services: (a) Location-sensitive billing; (b) Automated advertising; (c) Special announcement; (d) Mobile commerce security; (d) Enhanced call routing; and (e) Tolls and ticketing; (2) Information Services: (a) Mobile yellow pages; (b) Traffic reports; (c) Weather notifications; (d) Navigation information; (e) Wireless Internet services; (f) tourist services; (g) Dating and Games; and (h) Logistics management; (3) Tracking Services: (a) Commercial fleet management; (b) Find-a-friend, -pet, etc.; (c) Buddy service; (d) Asset tracking; (e) Public safety dispatching; and (f) Agency personnel safety; and (4) Assistance Services: (a) Emergency notification; (b) Roadside assistance; (c) Health- and medical-related location ID, and (d) Efficiency enhancements for business applications.
In Japan and Korea, nearly 100 high-precision location based services are available for approximately 3 million gpsOne terminals. Remember Harry Potter’s Marauder’s Map? In July, KDDI introduced GPS MAP, which allows users to see in real-time the location of all personnel or vehicles with compatible handsets, which opens up many potential uses in sales, delivery, maintenance, nursing, healthcare, security, and event management. Companies use a PC and an Internet connection to incorporate central push functions to make requests or obtain real-time location information. GPS MAP displays the movement histories of commercial vehicles, but handsets are equipped to maintain privacy when not working. For consumers, finding a friend or a blind date at the mall, sports arena, casino, or bar becomes very easy: just check the Friend’s Map and send a photo e-mail or make a call. By combining positions, maps, and voice services, carriers report increased wireless data use and voice traffic as consumers interact with each other using location information and these moving high-resolution maps. Verizon Wireless is offering the Vindigo City Application, which, at $3.75 month, provides concierge services by using map-and-directions-based information on restaurants, museums, and entertainment for 27 major cities.
Also in July, SnapTrack and TeleCommunication Systems announced an agreement to provide carriers with a shared service bureau carrier-class Position Determining Entity platform that combines SnapTrack technology with TCS’s Xypoint Location platform. This provides carriers with an alternative for smaller carriers to deploying their own E-911 system to increase safety and save lives. And, by using these established technology platforms and new service bureau, carriers can offer specialty consumer and enterprise location-based services now.
More Tipping Power. The FCC E-911 mandate means that Qualcomm will provide a position location coprocessor in all but entry handsets beginning with its 2003 6xxx series. As it becomes less of an option and more of a standard element, notice that buying the ASSP package itself automatically pulls through this new gpsOne technology. Economies of scale from using a standard ASSP platform drive this decision and guarantee the eventual ubiquity of the best solution, gpsOne. Of course, Qualcomm also includes position location because of its power to pull related data services offering compelling value. Thus, the low-end user who buys for voice becomes educated about the value of data services.
The Chasm Crossing Elevator Message: For operators who must meet the E-911 mandate, but who are dissatisfied with traditional GPS, SnapTrack’s technology-agnostic Assisted-GPS provides a fast, accurate position fix consistently, even indoors, unlike E-OTD, which requires new infrastructure for synchronization, or OTODA, which does not seem to work, Qualcomm has put together a cost effective and highly accurate system that guarantees privacy, meets the FCC mandate, and boosts ARPU.
The early booming success of position location services in Korea and Japan suggests that, in time, these services will become widely adopted globally. If operators, say, in India or Latin America do not offer the service originally to low-end subscribers, but add it later, then BREW will simply activate the coprocessor, if its present. With the augmenting development of diverse and appealing applications, position location itself adds perceived value because it is a tremendous resource, no, more, a set of multiple and fecund resources that will become still more essential priorities as gpsOne applications become increasingly available, cheaper, convenient, and easy to use. A new day in a global world is coming, a world of connections, always on, always with you, precisely pinpointing where you and your friends are now or where you wish to be.
From the perspective of strategic architectural control, the basic ASSP platform has expanded through the acquisition of another breakthrough technology architecture. SnapTrack’s breakthrough idea was the innovative distribution of functions among intelligent thin-client, intelligent-server-software-as-codified-knowledge, and intelligent wireless communication network to solve the problems of yield, speed, and accuracy in GPS position location while keeping a watchful eye on issues of personal privacy, spectral efficiency, conserving power, and reducing size. How could Qualcomm not snatch up this complete end-to-end solution to extend its platform into the FCC-required and service-promising domain of position location?
Besides, a new architecture that is integrated into a platform of architectures does more than create competitive advantage; it also creates sustainable competitive advantage. The extent of its sustainability is a function of how well it continues to satisfy compelling needs into the future. Position location is destined to become a required feature for modern living, offering a connection to both increased safety and intimacy. Thus, not only does the basic platform extend its scope with position location, but also, when it joins cdma2000 1X and 1xEV-DO, the combined benefits scale, amplifying two dimensions of power: power-as-mastery and power-as-added value. Both depth and scope, both richness and reach, are scaled up because power-as-mastery created a scaling up of power-as-added value. [And, yes, I do intend the implicit contrast of power-as-mastery with the dominating use of power-over the value chain, which the Gorilla Game appeared to prioritize.]
This combination of architectures in the expanded platform invites application developers to exploit this new opportunity for business and consumer services. Such platform-synergies, thus, become a scale-dynamic that invites, nay, seduces further offscale-dynamic network effects from both software developers and subscribers. Safety, state, condition, circumstance, and whereabouts of family and friends are a constant human concern. How many of us have heard, “If you’re not with me, I want to know where you are to be sure that you’re O.K.,” or, simply put, “let’s keep in touch.” Intimate connection is a deep human need, always on, always with you.
So, the contrast with Europe’s 3G solutions becomes starker. As I write this, Nokia has just delayed, with some face saving waffling, sampling of GSM/GPRS/UMTS from September to “early” in 2003. It falls further behind the MSM6200, which sampled three months early in mid-June. The 6200 offers a complete solution to GSM and UMTS network compatibility. It includes the [future] possibility of a synchronized UMTS network, but still meets the letter, if not the anti-competitive spirit, of European WCDMA standards. Plus, the 6200 adds a superior system of synchronized position location. Works better, is cheaper. The long-awaited and still-undelivered GSM/UMTS handset from Nokia cannot include a set of position location coprocessors that work across these European networks. Accurate position location, easy roaming, and reduced costs are significant competitive advantages.
By adding position location and multimedia to the equation, Qualcomm’s added value compounds, but Nokia’s added value correspondingly diminishes. Sounds to me like a manifestation of the reciprocal laws of increasing returns and decreasing returns. And, it is just the beginning of those reciprocal effects given the additional breakthroughs in data optimization and ZIF-architected world phones. When MSM6300 overlays cdma2000 1X on GSM’s present service system by introducing the eagerly awaited GSM1x phone in 2003, this becomes a no-brainer for Asia because it both increases voice capacity and drives data ARPU by providing unequaled positions location, multimedia services, and BREW’s enriching applications.
Given that European operators are deeply in debt to their own governments for 3G spectrums, and given that the Telecom Bubble has burst, the pressure to save your job, your company, and your national heritage mounts. This inability to roam in asynchronous GMS/GPRS/UMTS networks across multiple modes, bands, and networks and the additional inability to use advanced position location technology and services is a decided handicap, perhaps, a decisive one that may well tip the market toward Qualcomm’s competing multimode 6200, or to its GSM1x in the MSM6300, or with 1xEV-DO as in the MSM65000, or the complete worldwide radioOne solution that follows in the MSM6600. The business of business is to stay in business. A single defection, and European Solidarity has no clothes. As the wag might say, “Bullshit talks, but performance walks.” Qualcomm walks its talk. |
