January 2004

SIP For Wireless IP Applications

BY JONATHAN ROSENBERG, Ph.D

Yes, I have to admit it. I have a problem. I suffer from Ambulatory Communicator Covetousness (ACC). This disease, which strikes mainly technology professionals, causes the sufferer to feel intense jealously when they see someone else with a mobile phone that is cooler, sleeker, faster, and otherwise better than the phone they currently own. There is no cure, and the only treatment -- phone trade-ins -- provides only temporary (and expensive) relief.

Why do so many of us suffer from this disease? In the field of telecommunications, wireless is where the innovation is happening. It is there that we are seeing new features and services to improve productivity. Sometimes, these new features are on the phone itself (cameras, PDA capabilities, and so on), and sometimes they are provided by the operator. There, new features and services such as Push-to-Talk, multimedia messaging, location-based services, and presence are being rolled out by operators and being admired by consumers, like myself, who suffer from ACC.
In the telecommunications field, we usually associate innovative and cool applications with the Session Initiation Protocol (SIP). However, most people don�t normally associate SIP with the cool new wireless services and applications that are now becoming available. Most folks associate SIP with wireline services, or with 3G wireless, which has yet to arrive -- if it ever does.

However, this is not true at all! SIP is playing an increasing role in the emerging wireless applications that are being rolled out today. These applications generally run over the so-called 2.5G data networks -- CDMA 1xRTT, supported by Verizon Wireless and Sprint PCS, amongst others, or GPRS, supported by T-Mobile and AT&T Wireless, for example. These networks provide dialup-speed access (20-50 Kbps). Carriers have spent a lot of money deploying the infrastructure to support this service, and are anxious to roll out services that make use of it. Some of the most exciting services are Push-To-Talk, MMS, and location-based services. SIP plays a key role in many of these services, especially Push-to-Talk.

Push-To-Talk (PTT) has been successfully marketed in the United States under the DirectConnect brand by Nextel. It provides a walkie-talkie type of experience, allowing users to press a button, and instantly be connected to the recipient. Each side can talk in turn. Nextel has a proprietary network, called iDEN that supports this service. Given the success Nextel has had with the feature, other U.S. carriers are now in the process of deploying PTT on their new CDMA 1xRTT networks. Verizon has already launched and Sprint PCS is following shortly.

So, what role does SIP play in PTT? It plays two roles. Firstly, PTT is a VoIP application. The voice is carried over the IP network. It�s not as interactive as traditional VoIP (and therefore can work over the 2.5G networks), but it�s VoIP nonetheless. As a result, several carriers are using SIP to signal and manage the PTT sessions. Generally, the SIP messages over the air are compressed in order to save bandwidth, as the air interface is a precious resource. The Internet Engineering Task Force (IETF) has even developed standards for how to compress SIP -- a technology called SigComp, which is specified in RFC3320. Aside from the compression, SIP�s usage for PTT is not all that different than how it is used to signal a basic VoIP call. Unlike a basic VoIP call, a PTT call is answered automatically, and is accompanied by floor control signaling to ensure only one party can talk at a time. There are no standards in place yet for this floor control, but work on this is in progress within the newly-formed Centralized Conferencing (XCON) working group in the IETF to develop them.

The second way that SIP plays a role in PTT is through presence. Although presence is often considered in conjunction with Instant Messaging (IM), presence is actually quite separate. Presence represents your ability, willingness, and desire to communicate using a variety of different devices, communications means, and media types. My presence is affected by many sources. Whether or not my cell phone is on, whether or not it is in a call, whether or not Outlook says I am in a meeting, whether or not my IM application on my PC is connected, and whether or not I am in a call on my enterprise PBX -- all of these affect my presence.

What has this got to do with PTT? Everything! Firstly, PTT itself provides another source for presence. Whether or not I am in a PTT call says something about my availability for voice calls and other PTT calls. Perhaps more importantly, presence is a natural capability to add to PTT. When users enter their PTT address books, icons next to the entries can give an indication of whether that user is available for a PTT call. Knowing this ahead of time is very valuable for PTT as it is a very disruptive form of communications. When I Push-to-Talk a colleague, anything I say comes out of the speaker on their phone without any explicit acceptance from them. If that colleague is in a meeting or cannot be disturbed, the result is a poor user experience. In such a case, my colleague can set their presence to �unavailable.� When I enter my address book, I can see this, and therefore not bother them. Presence also allows users to avoid failed calls. After pressing the talk button, it can take the network anywhere from one to several seconds to establish the call. If the called party is not available (because their phone is off, for example), the caller receives a buzz to indicate that the call could not complete. These buzzes are an annoyance, and they degrade the overall user experience. With presence, a caller can know ahead of time that the target user is not available, and therefore not try to call them at all.

Presence also enables �opportunistic Push-to-Talk.� Opportunistic PTT is a specific example of opportunistic communications. Opportunistic communications are any communications that take place because the caller knows ahead of time that the call will succeed, and therefore the caller makes the call. If they had not known a priori of the availability of the called party, the call would not have been made. Today, opportunistic calls are facilitated by e-mail exchanges or by a human assistant. But, presence makes opportunistic calling much more likely. When a user enters their PTT address book, they can see, because of presence, whether the people in the list are ready for a call. As the user scrolls down the list to (for example) Bob�s entry, they may say to themselves, �That�s right -- I needed to talk to Bob about that customer contract�. If the presence icon in the address book indicates that Bob is available for a call, the user is going to be more inclined to make the PTT call. Why? Because they know that the call is likely to succeed.

Indeed, an analogy can be drawn to �cube shouting.� This frequent form of communications occurs between office workers that sit in neighboring cubes. When one worker (Bob) wants to ask another worker (Judy) a question, Bob has several options. He could pick up the phone, he could send an e-mail, or he could IM. Those, however, take too long for simple questions with yes or no answers. In these cases, the most efficient way to get his question answered is to just yell it out -- �Hey Judy, do you have those sales figures?� If Judy is around, she can answer by yelling back.

When it works, cube shouting is fast, efficient, and the best choice for certain forms of communications. But, it can only work when two conditions exist. First, when the shouter knows that the shoutee sits nearby, and secondly when the shouter knows that the shoutee is actually at their desk. PTT brings a technology solution for cube shouting, and allows the notion of �nearby� to apply to anyone with a PTT-enabled phone. For teams of office workers, this is a huge benefit. Teams that work closely together no longer need to sit close to each other in order to benefit from the efficiency of cube shouting. However, the other condition for usage of cube shouting still needs to be met -- knowing that the shoutee is at their desk. That�s where presence comes in to play. It tells the shouter about the availability of a user at their virtual desk.

The conclusion is that presence is incredibly compelling as an enhancement to PTT. Indeed, without it, PTT cannot adequately provide an equivalent to cube shouting as a communications medium.

Back to the original question -- what does SIP have to do with this? The IETF has specified a set of extensions to SIP for presence and instant messaging. These extensions, known collectively as SIMPLE, have become widely adopted across the industry as the mechanism for enabling interoperable presence and IM. It has been strongly embraced by the wireless industry, as evidenced by its adoption by the 3GPP and 3GPP standards bodies. Several of the PTT applications getting deployed today make use of SIMPLE in order to add presence capabilities.

So, next time any of you fellow ACC sufferers see someone with one of those new-fangled PTT phones, that pang of jealously can be accompanied with a sense of awareness that, behind the scenes, SIP is at the heart of this particular bit of coolness.

Jonathan Rosenberg is Chief Technical Officer at dynamicsoft, Inc. Leveraging acknowledged leadership in wireless and Internet standards bodies, dynamicsoft enables subscriber-aware services today while increasing the long-term value of our customers� networks.

If you are interested in purchasing reprints of this article (in either print or HTML format), please visit Reprint Management Services online or contact a representative via e-mail at [email protected] or by phone at 800-290-5460.

[ Return To The January 2004 Table Of Contents ]