SIP: The key to new services

The IP telephone revolution has been a strange one for both buyers and end users: On the one hand, a large technology switch has been taking place over the past eight years; it is impossible to ignore the fact that IP PBX shipments not only have larger growth, but have actually outstripped legacy PBX sales within only the past two years. There was a time not long ago when most attendees at the industry standard pulver.com "Voice on the Net" conference quite openly believed that VoIP was all nonsense and hype. Inside the industry, even the ones who still don't "get it" know it's here.

And yet, despite the unmistakable changes readers of Telephony have been seeing, it might well prove to be the most invisible technology switch ever, from the point of view of the buyer and the end user. People who are buying want PBXs--that's it. End users just want to be able to pick up the telephone and dial--they really don't care about any new features or services. I walked into a government office recently and was surprised to see the receptionist using a brand new Cisco 7960 IP phone. When I asked her what she thought of the new phone, her response was very revealing: "Huh?"

So what happened to the promise of a new world of telephone services that IP telephony would bring? Will there really be no benefits to the end user as a result of the largest technology switch in forty years, since touch-tone phones replaced rotary dialing? The truth is that the promise is not at all dead, and new things are on the way, despite the apparent telecom doldrums and despite the current taste for "nothing new." Services that fulfill the promise may be unexpected, and they may even be "invisible"--e.g., the faster call setup time that occurred with the introduction of SS7. But they can start today, with the equipment users have today on their desktops.  This article describes how SIP signaling makes that possible.

What is a service anyway?

Telephone calls are made up of two pieces: signaling and media transport (bearer). Media transport is easier to define--everyone can agree on what media is (especially voice media) and on what transport means. Telephone signaling is a bit harder to define--I define it as "the transportation of small amounts of data between or among a set of entities for the purpose of enabling or facilitating media transport between or among endpoints." The essential thing about signaling is that it means to transport bits of data so that people can exchange media--for example, there are signals that emanate from my telephone so that your telephone will ring; there are signals that emanate from inside the phone system so that your telephone will show my phone number as caller ID; there are signals that move entirely inside the telephone system so that a proper bill can be created for the call. In every case, signaling involves passing bits of data around that are required if people are to be able to talk on the phone.

Essentially every telephone 'service' is a result of some extra signaling which is appended to a telephone call, to make it more useful and valuable to the end users. The service that is valuable is in the extra data accompanying the call. This data might include actual voice recording files (as in voice-mail or prompts), or other media files (such as small pictures that flash when a call comes in), but more often than not it is pure signaling data--such as the special tones to identify call waiting, or the digits of caller ID, etc.

Once you understand this, you can understand why the IP telephone revolution in its current phase hasn't led to any new services. All the vendors are occupied with creating standard look-and-feel handsets that happen to have an RJ-45 plug rather than an RJ-11 plug. Because they look and feel exactly like the handset you had yesterday, there isn't any user interface element that could enable you to have a new service.

This is the essence of the coming revolution in telephone services--how will users dial their telephone calls in the future? What extra data will accompany the calls to make them more valuable? What sort of equipment will the end user need to be able to use these new services?  Anyone who can really figure out what end users will want to augment their calls, will certainly be able to find the engineering support to signal these new kinds of services.

SIP: The bearer of next-gen services?

"Just because you're paranoid doesn't mean they're not out to get you." How true that platitude rings in our terror-laden world, and similarly: "Just because there's huge amounts of hype around SIP doesn't mean it's not the biggest revolution in telephone signaling since touch-tones." The world has seen so many new signaling protocols over the past 20 years, from ISDN to the ITU-T's H-series to the current instant messaging and presence wars, that it's hard to take any of them too seriously, or to believe that the protocol could possibly make a real difference. And yet, there are reasons why SIP is truly different:

1. SIP is not just a voice telephony protocol. It is more of an overarching protocol that enables people to find each other and together decide very quickly how they want to communicate. When your telephone rings, you know someone wants to talk to you. With modern cell-phones, it's possible that they want to send you a text message or even a picture. SIP enables people with IP-connected devices to 'ring' each other for essentially any kind of mixture of real-time communications--voice, video, text, image, document, Web page, etc. Simple versions of these services are front-and-center in today's mobile phone marketing wars. In the future, richer combinations of voice, text, image and documents will be ordinary phone calls, and only SIP, among all the signaling protocol standards available today, can enable such calls.

2. SIP is designed from the beginning for groups of people, not just for two-party calls. Think of today's global email infrastructure, and ask yourself what it would look like if every time you wanted to write an email to more than one remote party, you had to stop your current email and switch over to a "conference email application." Again, SIP is unique among standard telephony protocols in its ability to transport signaling among groups. If the telephone is ever to become the basis of a useful groupware communication service, SIP will be at the center of its signaling.

The single most important thing about SIP is that it's pure signaling--SIP has nothing to do with RTP and voice codecs, and could just as easily be used to signal a wide range of communication services that have no voice element at all. Think of SIP as the "next-generation way to ring the telephones of others"--whatever "ring" happens to mean in the future, and whatever "telephone" happens to mean in the future.

Carriers, enterprises, end users: who benefits from new services?

In order to gain wide acceptance, a telephone service must be beneficial for three groups: the operators who enable the services, the end users who use the service, and the entity that pays for the service (often but not always the enterprise that employed the end user). The single biggest question facing the deployment of new services is this: Is this old rule still valid?

As long as the only advantage of IP telephony is that it is "cheap," end users are not going to embrace something new, carriers are not going to do anything to offer the new services, and those who pay the bills will be motivated to do nothing but reduce the bills they pay. It's also true that IP-handset makers will be motivated to reduce their COGS, not figure out the new user interface to dialing. Right now, it's mobile handset makers who are at the forefront of user interface innovation as a driver for service innovation. The simple addition of user-controlled ring-tones has generated tens of millions of dollars in new service revenue over the past five years.

This points to another aspect of the current rollout of new services: essentially all of the successful new services don't really involve IP--they include text messaging, photographic caller ID, new kinds of prepaid cards, etc. But they absolutely fit into the model I have proposed here: They all depend on new telephone signaling, on data that accompanies and augments the call; they all bring tangible benefit to the carriers, to the end users, and to the entities that pay for the call. As more and more desktops and mobile devices are IP-enabled, more and more of the telephone calls can be "IP-enriched"--whether or not the call involves VoIP.

How will we start the IP telephony services revolution?

The conclusion you might draw from this could seem bleak: that we will have to wait through at least one more generation of handset equipment before we can begin to see really new telephone services coming out in America. However, we at eDial see a way out of this impasse, based on the following observation: The end users who could benefit from new telephone services are also Internet-connected near their desktop phone. Their desktop user interfaces offer them rich possibilities to create and access signaling data to accompany and enrich their phone calls. There is already a wealth of data inside their enterprise directories and e-mail/IM communications systems that they could use to make their phone calls more valuable. What is missing is the signaling infrastructure and the seamless user interfaces which will allow them to tie it all together. Here is a short list of services that would use desktop or mobile applications in an essential way:

1. Dialing phone calls directly from directory entries

2. Getting groups of people on the telephone quickly and painlessly

3. Augmenting telephone calls with instant messaging as "interactive Caller ID" or as a sidebar chat during group calls and conference calls

4. Leveraging presence information as "smart lights" on the telephone

5. Quick sharing of pictures, files and Web pages during the call

Note that every one of those services makes perfect sense with ordinary phones, just as well as with IP-bearer phones. These services have nothing to do with voice over IP--they are entirely enabled by signaling over IP.

Each one of these services leverages data and applications that already exist to make the telephone call better, faster and more productive--and hence ultimately more valuable for the end users, the carriers and the enterprises that will pay for them. In fact, each is available today, using a combination of existing telephones, web browsers and SIP-enabled telephony servers. SIP is the technical protocol that enables the signaling infrastructure, which will make these telephone services possible. Now that the industry is absorbing SIP, we are ready for the next step--to create the user experiences that will turn technology into services.

Scott Petrack is CTO for eDial Inc.

Visit eDial online.