The Interface between Next-Generation
Application Servers and Media Servers: SIP and VoiceXML
(Continued from Part 1)
SIP and VoiceXML Enable New Functionalities and Services
SIP, coupled with VoiceXML, drives real-time media processing in a packet network. This is helpful when building enhanced services such as voicemail, conferencing or unified communications. In addition, SIP and VoiceXML have introduced many new functionalities that were not previously possible. With SIP and VoiceXML for example, the application need not be completely re-written for each type of user interface. A re-usable, interface independent, service logic makes application development much quicker and easier. Service logic is made up of user registration and pre-qualification commands such as, 'say hello', 'say give me your calling card number', 'start collecting digits', 'Do I have enough digits?' These commands are used to pre-qualify a user to access an application.
Prior to SIP and VoiceXML, this service logic had to be embedded in the application. It normally reads as follows:
- line 1: play a prompt
- line 2: did the prompt play?
- line 3: collect digits
- line 4: get the first digit
- line 5: get another digit
- line 6: get another digit, etc.
The application, however, does not need to be bogged down with these instructions that don't directly relate to its core functionality. In a calling card application, for example, the application just needs to have a valid card number to operate.
Before SIP and VoiceXML, the service logic had to be programmed from the beginning along with the application itself for each user interface that the application was ported to such as DTMF, voice response, HTML, or WAP. SIP and VoiceXML work with a servlet, a small server application that enables this interface portability. With this servlet, the application developer can port their application to the user interface of their choice and the application will stay unaltered no matter which interface they chose. This makes it possible to write the application part once and seamlessly port it to any user interface, saving a great deal of time in the development process.
The off-loading of the user interface from the application to VoiceXML enables server efficiencies that make the network scale better, leading to operational efficiencies. This also significantly cuts down on the time that it takes to install media servers and to deploy applications in a production environment.
Implementation Challenges
Detractors may say that its openness can be a drawback of SIP, whereas with MGCP and H.248 standards have been already established. These standards are restrictive, however, and most applications written using these protocols sound very similar. With the SIP and VoiceXML development model, however, there is much more room for innovation. SIP proponents are addressing this by putting forward a set of recommended best practices to make it easier for new comers to get started without limiting the freedom of more advanced users.
Examples of these best practices include draft proposals submitted to the IETF for how to have a media server play announcements using SIP alone or for how to reference VoiceXML in a SIP INVITE message. A SIP INVITE message invites a party into a call or invokes a media server interaction with the caller and the called party to do such things as, play an announcement, bridge a person into a conference call, run a VoiceXML script or to perform transcoding.
Conclusions
SIP & VoiceXML are already familiar to a much wider base of programmers than more proprietary, PSTN-oriented, standards. The service development time is also much shorter, using SIP and VoiceXML. VoiceXML is based on open web-based programming languages such as XML and HTML and SIP on open Internet protocols such as HTTP, for which there is a much larger set of support tools and development packages.
SIP and VoiceXML make real-time media processing possible and affordable in a packet network. They enable much quicker, more cost-effective application development and deployment. They provide a broader range of tools and access to a much wider pool or programmers. SIP and VoiceXML will foster innovation in IP enhanced services as no other protocol has to date.
back to the top
Copyright © 2001 VoiceXML Forum. All rights reserved.
The VoiceXML Forum is a program of the
IEEE Industry Standards and Technology Organization (IEEE-ISTO).
|
