The WAP Wave
What will WAP bring? A rising tide of WAP services has appeared since WAP 1.1-enabled clients rolled out in parts of Europe in the fall of 1999, and will swell even further with the ramp-up of WAP-enabled phones in the United States this year. The deployed and envisioned applications cover a wide spectrum: wireless data services for the mobile office (two-way email, unified messaging, fax, notification services); “localized content” for weather, directions, news and traffic; mobile-to-mobile instant messaging that has already demonstrated rapid growth in both Europe and Japan; games and entertainment that have had great appeal with the youth segment in wireless markets; and many third-party applications taking Internet e-commerce into wireless (travel, stock trading, banking, etc.). And, yes, you can buy a Coke from WAP-enabled Coke machines in Finland, a showcase country for WAP in which 67 percent of the population uses mobile phones and a youth culture has embraced wireless technology.
According to WAP proponents, this array of services will continue to blossom (and spread to the U.S. market), propelled by the youth market, early adopters, business needs for the mobile office, traveler aids, and the general information addiction of the Internet generation.
The momentum for mobile access to the Internet will undoubtedly continue to grow for wireless service providers. May 2000 saw the announcement that for the first time the number of wireless subscribers connecting to the Internet surpassed the number using fixed networks. It is interesting to note that these statistics are driven largely by NTT DoCoMo’s hugely successful iMode service, rolled out in early 1999 in Japan. While many claim that the special marketing advantages enjoyed by iMode in Japan will not translate to success elsewhere, WAP technology will likely have to prove itself in direct competition: In May 2000 DoCoMo invested 5 billion euros in the Dutch firm KPN NV’s KPN Mobile unit and, separately, acquired 20 percent of the biggest U.S. GSM operator, Voicestream. Many analysts speculate that one of the reasons for these investments is to bring iMode-like services to the European and U.S. markets. In any case, iMode is a true mass-market mobile data success story—running at a distinctly unsexy data rate of 9.6 Kbps. (See sidebar, “If WAP is here, is the game over?”)
What Problems Does WAP Solve?
What infrastructure is necessary for the services described above? Or, to restate, what are the problems that the WAP Forum felt needed solving in order to enable Internet services for mobile handsets?
First is the challenge of delivering content effectively over air interfaces. Wireless transports generally have quite limited bandwidth and high delays in comparison to the wired Internet. As a result, the current WWW content (image-rich HTML pages) and its transport (HTTP/TCP/IP) are not optimized for delivery over wireless links. Another challenge—one that mainstream TCP/IP solutions can’t meet—is robustness in the face of varying QoS and intermittent drop-outs and reconnects.
The WAP specifications are independent of the underlying bearer service, and interfaces between its protocol stack and many air link technologies have been defined to date as part of the WAP standards. Current WAP service trials and early deployments are using existing low-bandwidth wireless data networks, which support transmission at rates as low as 9.6 or 14.4 Kbps. Even GSM networks transitioning to packet-based GPRS (General Packet Radio Service) are today generally supporting only modest bandwidths of 20–60 Kbps, much less than the ultimate GPRS/EDGE solution that enables full-duplex 384 Kbps transmissions between the mobile client and its base station. Thus, the transport optimizations of the WAP stack are much needed to enable effective service deployment in most wireless networks today.
The WAP Forum also addresses the challenges of wireless content development. For example, mobile phones and other handheld computing devices are characterized by small screens, limited memory, difficult text entry and slow processors. This environment will not support bulky HTML markup, image-rich pages, heavy use of scripts, large forms, and other features common on the WWW. The WAP solution, the Wireless Application Environment (WAE), is to reengineer content authoring with a set of platform-neutral markup standards.
Carriers, Content Brokers and the Walled Garden
How will WAP-enabled WWW services get to the user’s handset?
Figure 1 / pg. 38 shows the architecture for wireless WWW-based data services. Wireless carriers are taking two approaches to obtaining content. For key portals and high-margin content providers (Yahoo!, Cnet.com, Travelocity.com, Fidelity Investments, etc.), service providers are going directly to the content site and striking custom deals whereby content providers pay for premium space on the wireless device. For a wide array of content, however, carriers are turning to content brokers—a new class of companies that aggregate and organize Internet content from a wide range of sources. Companies such as InfoSpace, Screaming Media, Giant Bear, Elocal and their partners typically perform several tasks as part of content aggregation for the wireless customer: enabling smaller content sites for WAP, organizing and managing real-time feeds, collecting local content and organizing it (often by zip code), and offering a portal where wireless users choose the Internet sites that will appear at the top level on their phone screens.
A key concept here is the “walled garden,” referring to the limited number of links that can be shown on the wireless client at one time. In an environment where user input is difficult, screen real estate is very limited and users themselves are in a hurry, a content provider must get a place on or near the top-level menu (inside the walled garden) or risk never being accessed by users. A Phone.com study, for example, found that access hits for icons fell off 50 percent for each menu level the user must traverse.
A crucial debate for carriers is how to use the walled garden to generate revenues. On the one hand, carriers could command top dollar for spots in the walled garden by not allowing users to change their icons. Although it is expected that most users will be satisfied with a few dozen of the top-flight Internet sites, some feel that users must have flexibility to personalize their site options—typically through a WWW portal operated by the carrier or the content broker. In any event, since many users will never reconfigure the defaults anyway, slots in the default walled garden have clear value.
Pricing Models and Billing Issues
How will carriers charge for wireless data services? The jury on revenue models is still out, and the current soup of wireless technologies, content delivery models, envisioned services and pricing plans will foster uncertainty for some time. The following are some of the pricing issues involved.
Advertising as a revenue base for wireless data
Carriers and content providers come together in the world of wireless data with very different perspectives on revenue models. While carriers have always focused on subscriber fees, many media companies and most large WWW content providers rely on advertising. In a recent quarter, for example, Yahoo made about 90 percent of its revenue from on-line advertising.
Will advertising-subsidized services be an important component of wireless revenue models? For some consumer market segments (e.g., price-conscious users), the answer is almost certainly yes. What would ads in the wireless handset look like? An interesting model being deployed by Yahoo! GeoCitiesGeoworks Corporation is personalized, perishable ads for wireless consumers. A user who opts for this service periodically gets, for example, an electronic discount coupon for a business in or near his hometown that advertises a limited-time offer. GeoCities reports that users like such ads because they are targeted and come from local businesses. Even so, “push” advertisers must be careful because users are very sensitive to anything that smacks of “spamming to the pocket.” Thus, ads have great potential, but the implementation of ad services will have to be carefully executed to avoid poisoning the well.
Carriers hold a powerful place in the value chain for targeted advertising: They possess information on the user’s location. In the ultimate scenario, fast food consumers will be pushed an electronic ad as they pass by a local franchise; or, upon request, a traveler will receive restaurant specials for the surrounding area. Location information in the carrier’s network is typically translated into zip codes at the access server, shown in Figure 1, in order to query servers at the content broker for localized content and advertising. Moreover, location information already has demonstrated value, because determining subscriber presence is the basis for the very successful mobile-to-mobile instant messaging services in Europe and Japan.
Flat fees to jump-start the market, but usage-based billing for future growth in premium services
U.S. carriers have adopted different approaches for the first round of plans that will hook customers on wireless data, but most are opting for simple pricing. Many (Sprint, AirTouch, Bell Atlantic) have offered data services via a free or low-cost recurring charge with additional per-minute access charges. Other wireless carriers—e.g., OmniSky and AT&T’s PocketNet—are experimenting with all-you-can-eat pricing, in recognition that flat-rate pricing has historically been effective at driving the adoption of new services. Moreover, U.S. consumers have become accustomed to flat rates for wired Internet services.
AT&T’s recently announced Digital PocketNet Service offers three levels of flat-rate service, with the basic service offering unlimited access to 40 fixed WWW sites for free. AT&T Kendra VanderMeulen, senior vice president of product development and strategy at AT&T Wireless, notes, “Consumers often need the free service inducement in order to take the plunge and sign up for wireless data. Once they experiment and discover value, many will choose to move up to our Plus and Premium plans.” The latter flat-rate options offer calendars, email, fax and general WWW browsing.
True premium services generally imply content-based pricing, or a mixed model that includes a per-session or per-transaction fee along with or in place of other charges. In WAP service trials (and for NTT’s DoCoMo iMode service), certain premium services such as instant messaging and high-value content downloads are priced per access. Services such as instant messaging require carrier investment to build out the back-end infrastructure for subscriber presence and location information. At least in Europe and Japan, the youth market has very much accepted such pay-as-you-go pricing. A recent Wall Street Journal article reports that the average Finnish teenager spent $867 last year on mobile phone services in addition to basic airtime fees. (How much of this spending was for data services is not entirely clear.) Figures released recently on NTT iMode services show subscribers spending an average of $15 per month downloading Internet content, which is charged by the packet.
Fine-grained usage-based billing will become technically easier as network providers move to IP-based networks like GPRS/EDGE, where sophisticated packet-capture technology developed for the Internet can be deployed in the mediation layers. According to Stan Fontanini, director of solutions marketing at Narus. and a member of the WAP Billing Experts Group, “The Semantic Traffic Analysis technology found in Narus mediation products offers a wireless carrier the ability to obtain session-level information directly from the network [see “Usage Collection and Analysis in an IP OSS,” Billing World, March 1999]. This information can be aggregated on a per-user basis to enable fine-grained usage-based billing that keeps carriers in the value chain for wireless e-commerce.”
One device, multiple bills
Another interesting innovation in billing is having multiple billable clients for services performed on a single physical device. BellSouth recently introduced mobile data service accounts where a user with a single handheld device is billed for personal Internet access, the user’s employer is billed for corporate two-way paging, and a third party (Fidelity Investments) is billed for stock trading access. This kind of “reverse convergent” bill will make consumers laden today with multiple mobile devices very happy indeed!
Billing Solutions: What Is Needed?
Billing solutions for the wireless service provider are not well established, given the early stages of the industry roll-out. In the case of WAP trials and deployments, billing is most often based on airtime, and the only usage information available is that collected in log files at the WAP gateway. In recognition of the immature state of billing architectures, the WAP Forum recently formed a Billing Experts Group (BEG), whose mission is to develop an architecture and methodology for with advanced billing capabilities.
Ideal billing solutions for wireless providers face a number of challenges:
Ability to accept raw usage information from multiple sources
One mission of the BEG is to optimize the gateway log file procedure to record and provide more granular information for billing. But gateway-only usage collection has clear limitations, and future billing solutions will benefit from or even require usage information from sources beyond the WAP gateway. Such sources include network mediation systems, application server logs and user profiles maintained by the provider for customizing information flows.
To illustrate this point, consider the difficulties in the following scenario. Your mobile subscriber decides to watch live coverage of the Stanley Cup game on his WAP terminal. He receives 3 minutes of live ESPN coverage before he moves to a location where the available bandwidth is insufficient for the video feed. The video is replaced with an Associated Press still photo of the last goal, with live radio commentary from his local radio station. He moves again to an even poorer coverage area, loses the radio commentary and instead gets scrolling text commentary from, say, Reuters. A minute later he moves back into excellent coverage and gets back his video feed.
There is a clear need for a very sophisticated mediation device that can record and interpret these multiple events in the course of a single data call and present them to a billing system in a format it understands.
Flexibility for a diverse application mix
WAP content providers are already rolling out a wide range of services, and the set of services that providers will want to bring into their walled gardens will be very broad, since new services will be a key differentiator in the marketplace. Flexibility in pricing new services, adapting billing for users who want to reconfigure their service mix, and nimbly supporting rapid roll-outs for new services are all major challenge for billing system vendors.
Real-time support for prepayment
One immediate concern of the WAP BEG has been support for prepaid services. In Europe, prepayment for mobile voice has been very popular, especially among young users. Teens and twenty-somethings are also adopting WAP data services, especially the short message services (SMS) that deliver short text messages from one mobile handset to another. In the United Kingdom, providers have recently reported a doubling of SMS traffic each month, and total traffic now exceeds 1 billion messages per month. Thus, support for prepaid accounts is seen as crucial for these markets. Many pundits expect both prepaid services and SMS to play a much more prominent role in the United States in the next 12–18 months, and this would further fuel the growth of WAP. BT Cellnet is probably the leading proponent of WAP-based services in Europe, and it has specifically targeted a substantial portion of its limited stocks of WAP-enabled handsets in this segment.
Prepayment places requirements for real-time authentication and accounting on the billing and OSS infrastructure in order to authorize service for a give user. It also requires real-time monitoring and intervention when a user exceeds the prepaid limit. A flexible, modular software architecture will be required: Billing system components handling the real-time functions are pushed “out to the network gateway” and optimize the responsiveness of communication with back-end accounting systems. These “front-end” elements also facilitate real-time provisioning of services and fraud protection.
Other OSS Challenges
WAP and next-generation wireless services bring many other challenges to service provider operations support systems (OSS):
?Bandwidth management—As the bandwidth from the mobile client to the provider increases, service providers will have to become more concerned about the impact of individual users on the network. At 14.4 Kbps, one or two users have a limited impact. At 384 Kbps, a few aggressive users can inject large amounts of data into the core network, and digital audiovisual streams (likely candidates for such heavy usage) do not allow the aggregate network service to degrade gracefully.
?User Device Management—Many different types of devices will be used to access the network, not just phones. Being able to tie multiple devices to one account, as well as manage and take advantage of different capabilities offered by those devices, will have a significant impact on the OSS for provisioning, roaming and CRM. Carriers will need to profile not only customers and services but the devices that are employed within the network.
?Real-time provisioning—Service activation for mobile clients today is still largely manual, and automated activation is widely recognized as an issue. The WAP Forum is working on a provisioning architecture that emphasizes automation through communication between bootstrap code in the mobile device and trusted proxies in the WAP provider network. This work, however, is still in an early phase. To provision services, carriers will operate in multiple dimensions—with trading partners, content providers, resellers and other service providers. Provisioning a service in this environment means close coordination between all the parties involved. Each service provider will need to extend its systems to its partners.
CPP for Data
While Calling Party Pays (CCP) has not been widely deployed in the United States for voice, CPP for data is another thing. CPP for voice has been unsuccessful for various reasons: intercarrier billing, consensus on announcements, user acceptance and regulatory issues. For data services, users will certainly not want to be charged for ads to the pocket; and other opt-in push services will find CPP useful or in some cases even necessary.
Voice/data system integration
Many challenges will arise from the convergence of voice and data services. At many carriers today, voice and data are handled on separate networks and by separate organizations, suggesting at least the potential for rough spots in back-end integration. As application support like the Wireless Telephone Application (WTA) interface for WAP (and voice recognition for microbrowsers) tightly intertwines voice and data services, operational support for voice and data in the same user session will introduce new OSS requirements.
Consider a WAP application where a user is browsing content in Wireless Markup Language (WML) and then evokes a call from a WTA interface within the application. Is this sequence of events handled (and billed) as a multiple-mode session on the data side, or as two separate network sessions, one voice and one data?
Roaming
Roaming for data services is an open issue. OSS complications include interoperability (e.g., enabling premium location-specific services while roaming), privacy related to sharing user profiles and user location information, and settlement (are there quality of service discounts when premium services fail during roaming?). Even if carriers were to share a single air link standard, they would still use multiple infrastructure vendors whose data switches and OSS are not compatible, making roaming an enormously complex problem.
Conclusion
The opportunity in mobile wireless data is huge. Unlike other technologies that took a decade to roll out, the first-generation wireless data protocols are deployed, and rapidly evolving standards are nurturing a robust, capable infrastructure. Whether service providers can capitalize on this opportunity, however, hinges on two things: developing profitable pricing models, and cost-effectively building out a capable billing and OSS infrastructure.
At this point, the pricing and revenue models are not clear. Ultimately we will probably see a combination of flat rate for local services and airtime; innovative packages and bundles for target marketing; usage fees for premium content and services; and a variety of charge-backs for ads and third-party content.
With the uncertainty of pricing and the speed at which providers want to roll out services, the biggest challenge clearly lies in developing a flexible, effective OSS infrastructure. Here, the fundamental obstacle is to integrate mobile wireless data services with existing support systems. Bear in mind that this is no small effort because such systems are optimized for voice. Further, wholesale replacement of these legacy support systems is not an option, because such an approach is often simply not cost-effective.
Most likely, the wireless industry will migrate to component-based architectures targeted at retrofitting existing billing and OSS systems to incorporate wireless data. For example, already we see vendors offering specialized usage collection subsystems (e.g., Narus and XACCT) and rating systems (e.g., RateIntegration) optimized for wireless data services. Likewise, many expert integrators (e.g., Logica) can pull such best-of-breed systems together to not only support new IP-based network elements and innovative pricing models, but also leverage the provider’s existing investment in legacy billing systems. On the other hand, some carriers have chosen to take the low road by simply pricing services at a flat rate. In the end, of course, consumers don’t care how all this gets done. They just want their new wireless data services, now, at a reasonable price. Good luck.
Dr. Matthew Lucas is President and CEO of RateIntegration, Inc. (www.rateintegration.com) - a software development firm that builds stand-alone rating engines optimized for wireless, network, application and access providers. He is also a co-founder of the IPDR initiative and currently servers as Honorary Chairman. Dr. Lucas is a frequent contributor to Billing World, and is recognized as a leading expert in the field of IP billing, mediation and rating. He can be reached at mlucas@rateintegration.com.
Dr. Bert Dempsey is a technical consultant with TeleStrategies and an Assistant Professor of Information Science at the University of North Carolina at Chapel Hill. He also serves as a co-leader of the Internet2 DSI project. Bert can be reached at bert@telestrategies.com.
The authors wish to thank Chris Couch and Vincent Kavanagh at Logica for their expert input to this article.
All of these factors mean that the HTML/HTTP/TCP/IP solution for the wired Internet does not translate well into the wireless domain. As a result, the WAP Forum has defined its own wireless protocol stack of functions as a replacement for the Internet stack. The layers of the WAP solution are shown on the right side of Figure B, with the corresponding standard Internet solutions shown at the left. (The correlation is useful for understanding the role of each protocol, but their capabilities are not an exact match.) A WAP gateway contains both stacks, and it translates data from one to the other to enable communication between the wired and the wireless networks.
Consistent with standard protocol architectures, the WAP stack has a set of well-defined functional layers that are mixed and matched depending on the host environment. For example, a transaction-based service running on top of the WAP stack could interface directly to the Wireless Transaction Protocol (WTP) layer if the functions of the Wireless Session Protocol (WSP) layer were not relevant to it. Wireless Application Environment
To address the computing limitations of wireless devices, the WAP Forum has defined a platform-neutral standard for application development, the Wireless Application Environment (WAE). The WAE sits on top of the WAP stack inside a wireless client device. It is compatible with all major handset operating systems, including PalmOS, EPOC, Windows CE, FLEXOS, OS/9 and JavaOS.
Components of the WAE include Wireless Markup Language (WML), WMLScript and the Wireless Telephony Application (WTA) framework. WML is a stripped-down markup language based on a “deck of cards” paradigm for effective content presentation and easy navigation from screen to screen in the microbrowser on a WAP client. It supports (hyper)text-only content with very simple icons (1-bit resolution). WML is an XML-based language and a derivative of an early Phone.com solution, HDML (Handheld Data Markup Language), that is still supported in many microbrowsers.
To enhance the power of WML, WMLScript is a scripting language that enables dynamic actions using WML-based content, just as JavaScript and Jscript do for HTML-based WWW pages. However, WMLScript forgoes the advanced features of its wired WWW equivalents in order to reduce memory and processing requirements.
To support voice and data convergence, the Wireless Telephony Application (WTA) framework defines standards for embedding voice calls within data applications in an easily programmable way. The WTA will thus enable data-driven interface to directories, embedded help-desk numbers and similar applications.
Finally, though technically not a part of the WAE, an important application development mechanism in WAP is an explicitly defined architecture for “push” notification services.
Wireless Session Protocol
The WSP layer is the top of the WAP stack and as such provides the interface for WAE-compliant applications to access the WAP communications stack. In particular, WSP plays the role of HTTP in the wireless network—HTTP messages are carried in WSP using a custom binary encoding. At the WAP gateway, WSP-encoded messages are translated into standard HTTP messages for communication with servers, and in the reverse direction the server’s response must be encoded and carried by a WSP message. WSP also handles push functions for notification services to the wireless client.
Wireless Transaction Protocol
The WTP layer coordinates and optimizes transaction-based exchanges between the wireless client and the application server, such as for e-commerce. As with other protocols in the WAP stack, its emphasis is on reducing the number of packet exchanges and the amount of data in each packet in order to accomplish the task, in this case to complete a client-server transaction.
Wireless Transport Layer Security
The WTLS protocol handles secure communications, including authentication and encryption using standard schemes. It was developed to be more efficient for message exchange than current Internet standards, i.e., SSL/TLS (Secure Socket Layer and Transport Layer Security)or SSL (Secure Socket Layer). At the WAP gateway, secure information in the WTLS layer is converted into SSL/TLS/SSL format for transport across the wired Internet (and vice versa). This implies that the WAP gateway must be a secure platform in order to protect the end-to-end security mechanisms.
Wireless Datagram Protocol
The WDP is the lowest layer in the stack, supporting basic transmission of packetized information for upper layer protocols. For networks supporting IP, the role of the WDP layer is played by the standard UDP protocol defined in the TCP/IP stack. For networks where IP is not available, the WAP standard defines the functions that the WDP replacement must fulfill. Note that the WDP itself does not provide the strong reliability guarantees that TCP does. This functionality is not necessary for some applications and is handled by the session layer (or transaction layer) for others. TCP is the basis of most communication in the wired WWW, but was deemed too chatty and heavyweight to embed at the bottom of the WAP stack.
The bottom of the WDP layer is the adaptation layer, the glue that defines the interface to the underlying bearer service. The WDP adaptation layer is different for each bearer service, and the WAP standard already defines adaptation layers for a large number of existing air link technologies.
Is WAP the ultimate wireless Internet solution?
Many skeptics would say definitely not. Critics point out that WAP has a number of obvious weaknesses, including difficult data entry (try entering a URL on a phone keypad); a limited, text-based display that restricts the richness of the user experience; and a proprietary approach that breaks much of the content—and decouples WAP somewhat from the innovation curve—on the HTML/IP Internet.
On the latter point, it is important to note that, because WAP has so radically reworked the Internet space, delivery of general Internet content to a WAP device is not feasible, and services will be based on content developed for or adapted to WML.
Limited automated translation of HTML content to a WML format is a role of the WAP gateway, a communication server that sits between the wireless client and the wired Internet (see Figure A). However, in practice, most deployments are opting for either direct conversion to WML at the content server or, in some models, conversion at special-purpose proxy servers tightly aligned with the original WWW content server. While this “repurposing” of content into WML is enjoying much popularity, a key point is that WAP will not be an access solution for reaching the entire HTML-based WWW.
Critics of WAP summarize the situation by saying that, while the scope of the WAP deployment will create a de facto standard in the short run (and much content will be reformatted into WML), developing standards parallel to the existing Internet for a limited user experience will not be a compelling technology in the long run.
At the May What Month? Wireless Agenda 2000 conference held in Austin, Texas,What City? industry guru Andrew Seybold and his associates suggested that WAP would achieve only modest success, due to the limitations of small displays, difficult data entry and limited application capabilities. In this context they referred to WAP as “the Citizen’s Band radio of the Internet.” The label implies that WAP will play an important role by exposing users to wireless data services, but ultimately these users will want a richer experience than WAP can deliver.
The Seybold Group offered an alternative vision, at least for U.S. users, of mobile data services: wireless data on PDA-class devices using Bluetooth—an emerging (and widely accepted)[<
WAP proponents see their technology as a good first step for wireless that will evolve with network capabilities. They claim the WAP Forum work will converge with mainstream Internet/W3C standards in the long run, though how and whether this will occur is not altogether obvious. Many observers simply say that WAP is an interim solution with a limited lifetime of 2–3 years, and convergence may simply mean (re)adoption of current Internet and WWW standards into the WAP framework. (For example, Microsoft wants to see TCP as part of the WAP 2.0 protocol stack in order to support the reliable download of large multimedia objects.) This view is driven by the fact that major U.S. carriers are forecasting high-speed packet capabilities in some markets by the fourth quarter of 2001—144 or 384 Kbps GPRS/EDGE roll-outs—and national coverage by 2002–2003. In these networks, mobile clients will simply move to native Internet content (HTML/XML delivered by HTTP/TCP/IP) and WAP will fade. The WAP supporters counter that optimized content delivery will always be a need, and that even in higher bandwidth networks, as user demands grow the need for capacity relief will keep WAP relevant.
In any case, most vendors in the wireless space are hedging their bets. Most microbrowser solutions today support more than one mark-up language from the different versions of HDML (the pre-cursor to WML), WML, and HTML. Microsoft’s microbrowser, for example, supports both HTML and WML. Popular PDA solutions like the PalmPilot have not yet adopted the WAP approach, since these platforms are not as limited in hardware and software as mobile phones. The Palm Pilot VII, for example, uses HTML/Java technology for Palm-optimized WWW content presentation. OmniSky recently launched wireless WWW service (in partnership with AT&T Wireless) using the Palm client. This service offers access directly to standard WWW sites, some of which have a Palm-optimized version.
Meanwhile, on the server side, wireless portal packages like Oracle’s MobileOracle solution support customized WWW and database content translation for delivery to wireless devices, where WAP’s WML is one of the supported content filters.
In short, WAP is not the only horse in the race. While a plurality of wireless data solutions is inevitable in the short run, strong success by a competing technology (such as NTT’s I-Mode) could eventually reduce the WAP buzz significantly.
