heca van hauser

Sunday, May 27, 2007


Technical specifications
The bottom-most protocol in the suite is the WAP Datagram Protocol (WDP), which is an adaptation layer that makes every data network look a bit like UDP to the upper layers by providing unreliable transport of data with two 16-bit port numbers (origin and destination). WDP is considered by all the upper layers as one and the same protocol, which has several "technical realizations" on top of other "data bearers" such as SMS, USSD, etc. On native IP bearers such as GPRS, UMTS packet-radio service, or PPP on top of a circuit-switched data connection, WDP is in fact exactly UDP.
WTLS provides a public-key cryptography-based security mechanism similar to TLS. Its use is optional.
WTP provides transaction support (reliable request/response) that is adapted to the wireless world. WTP supports more effectively than TCP the problem of packet loss, which is common in 2G wireless technologies in most radio conditions, but is misinterpreted by TCP as network congestion.
Finally, WSP is best thought of on first approach as a compressed version of HTTP.
This protocol suite allows a terminal to emit requests that have an HTTP or HTTPS equivalent to a WAP "gateway"; the gateway translates requests into plain HTTP.

Wireless Application Environment (WAE)
In this space, application-specific markup languages are defined.

The primary language of the WAE is WML, the Wireless Markup Language, which has been designed from scratch for handheld devices with phone-specific features. WML is an XML-compliant format. However, since XML documents can take up a lot of room, a specific compression technique for XML documents was developed (wireless binary XML, or WBXML).

Maintenance and evolutions
The WAP Forum has consolidated (along with many other forums of the industry) into OMA (Open Mobile Alliance), which covers virtually everything in future development of wireless data services.

WAP 2.0
WAP 2.0 is a re-engineering of WAP using a cut-down version of XHTML with end-to-end HTTP (i.e., dropping the gateway and custom protocol suite used to communicate with it). A WAP gateway can be used in conjunction with WAP 2.0; however, in this scenario, it is used as a standard proxy server. The WAP gateway's role would then shift from one of translation to adding additional information to each request. This would be configured by the operator and could include telephone numbers, location, billing information, and handset information.

Some observers predict that this next-generation WAP will converge with, and be replaced by, true Web access to pocket devices. Whether this next generation (Wireless Internet Protocol to mobile) will still be referred to as WAP is yet to be decided. XHTML Mobile Profile (XHTML MP), the markup language defined in WAP 2.0, is made to work in mobile devices. It is a subset of XHTML and a superset of XHTML Basic. A version of cascading style sheets (CSS) called WAP CSS is supported by XHTML MP.

WAP Push
WAP Push, has been incorporated into the specification to allow WAP content to be pushed to the mobile handset with minimum user intervention. A WAP Push is basically a specially encoded message which includes a link to a WAP address. WAP Push is specified on top of WDP; as such, it can be delivered over any WDP-supported bearer, such as GPRS or SMS.

In most GSM networks there are a wide range of modified processors, however, GPRS activation from the network is not generally supported, so WAP Push messages have to be delivered on top of the SMS bearer. On receiving a WAP Push, a WAP 1.2 or later enabled handset will automatically give the user the option to access the WAP content.

The network entity that processes WAP Pushes and delivers them over an IP or SMS Bearer is known as a Push Proxy Gateway

Commercial status

Possible failure
WAP was hyped at the time of its introduction, leading users to expect WAP to have the performance of the Web. One telco's advertising showed a cartoon WAP user "surfing" through a Neuromancer-like "information space". In terms of speed, ease of use, appearance, and interoperability, the reality fell far short of expectations. This led to the wide usage of sardonic phrases such as "Worthless Application Protocol", "Wait And Pay", and so on.

Critics advanced several explanations for the early failure of WAP. Some are technical criticisms:

The idiosyncratic WML language, which cut users off from the true HTML Web, leaving only native WAP content and Web-to-WAP "proxified" content available to WAP users. However, others argue that technology at that stage would simply not have been able to give access to anything but custom-designed content.
Under-specification of terminal requirements. In the early WAP "standards", there were many optional features and under-specified requirements, which meant that compliant devices would not necessarily interoperate properly. This resulted in great variability in the actual behavior of phones. As an example, some phone models would not accept a page more than 1 Kb in size; others would downright crash. The user interface of devices was also underspecified: as an example, accesskeys (e.g., the ability to press '4' to access directly the fourth link in a list) were variously implemented depending on phone models (sometimes with the accesskey number automatically displayed by the browser next to the link, sometimes without it, and sometimes accesskeys were not implemented at all).
Constrained user interface capabilities. Terminals with small black and white screens and few buttons, as the early WAP terminals were, are not very apt at presenting a lot of information to their user, which compounded the other problems: one would have had to be extra careful in designing the user interface on such a resource-constrained device.
Lack of good authoring tools. The problems above might have been alleviated by a WML authoring tool that would have allowed content providers to easily publish content that would interoperate flawlessly with many models, adapting the pages presented to the User-Agent type. However, the development kits which existed did not provide such a general capability. Developing for the web was easy: with a text editor and a web browser, anybody could get started, thanks also to the forgiving nature of most desktop browser rendering engines. By contrast, the stringent requirements of the WML specifications, the variability in terminals, and the demands of testing on various wireless terminals, along with the lack of widely available desktop authoring and emulation tools, considerably lengthened the time required to complete most projects.
Other criticisms are oriented towards the wireless carriers' particular implementations of WAP:

Neglect of content providers. Some wireless carriers had assumed a "build it and they will come" strategy, meaning that they would just provide the transport of data as well as the terminals, and then wait for content providers to publish their services on the Internet and make their investment in WAP useful. However, content providers received little help or incentive to go through the complicated route of development. Others, notably in Japan (cf. below), had a more thorough dialogue with their content provider community, which was then replicated in modern, more successful WAP services such as i-mode in Japan or the Gallery service in France.
Lack of openness. Most wireless carriers sold their WAP services that were "open", in that they allowed users to reach any service expressed in WML and published on the Internet. However, they also made sure that the first page that clients accessed was their own "wireless portal", which they controlled very closely. Given the difficulty in typing up fully qualified URLs on a phone keyboard, most users would give up going "off portal"; by not letting third parties put their own entries on the operators' wireless portal, some contend that operators cut themselves off from a valuable opportunity. On the other hand, some operators argue that their customers would have wanted them to manage the experience and, on such a constrained device, avoid giving access to too many services.

Possible success
However, WAP has seen huge success in Japan. While the largest operator NTT DoCoMo has famously disdained WAP in favor of its in-house system i-mode, rival operators KDDI (au) and Vodafone Japan have both been successful with the WAP technology. In particular, J-Phone's Sha-Mail picture mail and Java (JSCL) services, as well as (au)'s chakuuta/chakumovie (ringtone song/ringtone movie) services are based on WAP. After being shadowed by the initial success of i-mode, the two smaller Japanese operators have been gaining market share from DoCoMo since spring 2001.

Korea is also leading the world in providing advanced WAP services. WAP on top of the CDMA2000 network has been proven to be the state of the art wireless data infrastructure.

According to the Mobile Data Association, June 2004 saw a considerable increase of 42% in its recorded number of WAP pages viewed compared with the same period in 2003. This took the total for the second quarter of 2004 to 4 billion.

Between 2003 and 2004, WAP made a stronger resurgence with the introduction of Wireless services (such as Vodafone Live!, T-Mobile T-Zones and other easily-accessible services). Operator revenues were generated by transfer of GPRS and UMTS data which is a different model to the Web, and usage was up. People are starting to use WAP and the early failures have been masked, as the real point of the system – access to wireless services and applications – has come to the forefront.

Spin-off technologies, such as MMS (Multimedia Messaging Service) (picture messaging), a combination of WAP and SMS, have further driven the protocol. An enhanced appreciation of device diversity, supported by the concomitant changes to WAP content to be more device-specific rather than being aimed at a lowest common denominator, has allowed for the content presented to be more compelling and usable. As a result, the adoption rate of WAP technology is on the upswing.

Protocol design lessons from WAP
There has been considerable discussion about whether the WAP protocol design was appropriate. Some have suggested that the bandwidth-sparing simple interface of Gopher would be a better match for mobile phones and Personal digital assistants (PDAs).

The initial design of WAP was specifically aimed at protocol independence across a range of different protocols (SMS, IP over PPP over a circuit switched bearer, IP over GPRS, etc). This has led to a protocol considerably more complex than an approach directly over IP might have caused.

Most controversial, especially for many from the IP side, was the design of WAP over IP. WAP's transmission layer protocol, WTP, uses its own retransmission mechanisms over UDP to attempt to solve the problem of TCP's inadequacy for high packet loss networks.

See also
Wireless transaction protocol
Wired Equivalent Privacy
Wikipedia access via WAP
Mobile development



Post a Comment

Subscribe to Post Comments [Atom]

<< Home