Sunday, July 24, 2005

Mobilisation and IMS...

Mobilisation and IMS...
By Paul Golding

There is still a lot of uncertainty about the future of mobile communications. This is easy to explain. Mobile telephony is nothing new. We already had telephones before mobiles and the transition is a very obvious step and mostly a matter of economics (i.e. making it cheap enough to do). However, everything else we are likely to do with mobiles in the future will be new. We tend to think of a progression or evolution from voice-based devices to "data" devices. However, there isn't necessarily a continuum. The future is about mobile computing, which is quite a different paradigm from mobile phoning.

Recently, I wrote a chapter "The Future of Mobile in the 3G Era" for the forthcoming title "Thumb Culture". I have been busy expressing the key themes in a presentation, mostly to provide a vehicle for explaining why, if at all, operators need to switch to using SIP-based signalling in their networks. In other words, "why do operators need IMS?"

Numerous whitepapers and sales presentations about IMS seem to concentrate on the topic of convergence, or seamless mobility. This usually means that you can use any wireless technology to make phone calls from the same phone. Outside, you would use 3G or GSM. Inside, you might use Bluetooth or WiFi.

What's the difference? It's cost. If you can make calls at home using a WiFi access point hanging off your broadband pipe, then the calls should be lower cost as you are paying for the broadband pipe already. Of course, this kind of idea works best if you have one phone number (your mobile number essentially), which works the same whether via GSM or WiFi. This is the promise of converged fixed/wireless services.

Where does IMS come into this? Well, to make a phone call, the phone has to exchange all kinds of signals with the core network that eventually connects one caller with another. A GSM phone only does "GSM signalling". The way signalling works includes how the signals are carried by the underlying radio interface, which is very well defined for GSM, as it is for CDMA and other standards. If we just call that interface "cellular" for a moment, then our mobile phones pass their signals using "GSM-over-cellular", which is all laid out in the GSM specifications. However, there's no such thing as "GSM-over-Bluetooth", or "GSM-over-WiFi" etc.

One company (Kineto) has come up with a "GSM-over-WiFi" system. They have done this using a clever trick. By pretending that your WiFi access point is a mini GSM base-station, your home becomes a cell in the cellular network. Hence, by adding a WiFi radio to the GSM phone and then using the "WiFi-as-cell" trick, they end up with a "GSM-over-WiFi" system, which looks like a "GSM-over-cellular" system.

This system, called UMA, means that your GSM phone will work on your WiFi access point. Instead of paying mobile call rates in the home, you will, in essence, get your calls from home "for free", as you're already paying for the WiFi-broadband bit. BT are about to offer such a service, called Bluephone (they orignally thought Bluetooth would be better than WiFi).

IMS offers an alternative. With IMS, there's a whole new way of doing all the signalling, and it's called Session Initiation Protocol (SIP). The beauty of SIP is that it is one of those wonderful protocols based on Internet Protocol (IP) that doesn't take much effort to understand nor to implement. In fact, there are plenty of free pieces of software out there that can implement SIP. It's the kind of thing a software undergrad could write quite easily. GSM, on the other hand, is not a barrel of laughs when it comes to implementation. Moreover, SIP, like so many IP-based protocols, is highly extensible, unlike so many non-IP protocols in the telecoms world.

SIP then, will run on any device that can connect with an IP network. These days, this includes a lot of devices. Even elevators can - and do - connect with IP networks. Certainly any wireless device, like a PDA or mobile, will have the required piece of software to "talk IP", often called an IP Stack. Naturally, our desktop and laptop PCs also talk IP and can definitely implement SIP.

Therefore, if the core of the mobile network is converted to SIP, instead of GSM (CDMA etc), then overnight the network can handle calls between any SIP-compatible devices, no matter how they implement the IP connection: over Cable, ADSL, WiFi, GPRS, Bluetooth, 3G etc. You can think of SIP like a Hotmail account. You can log-on from anywhere and then get your email. With SIP, you can log-on ("register") from anywhere and get your phone calls, voicemail etc.

From this explanation, it is easy to see why the focal point of SIP-based networks for cellular is often seamless mobility. However, the real power of SIP and IMS lies beyond seamless mobility.

In my opinion, 3G is IMS. What SIP enables is a generic "connecting" protocol. We don't necessarily have to be interested in connecting users to make a voice call. In other words, SIP isn't just a "calling" protocol.

Thus far, mobile networks have been almost exclusively concerned with voice calls. The data model that has emerged so far is WAP, which is essentially "Web-lite". The paradigm is connecting users with information in the form of Web pages (portals etc.)

However, the essential nature of mobile technology is connecting people. This Person-to-Person (P2P) nature will be a dominant feature of mobile computing. We need to grasp what P2P "connecting" is all about. Today, we talk to each other. But, tomorrow, we shall:


Click to play, to share, to view, to update, to invite, to compare, to tag, to consult, to message, to conference
…Click to connect!

GSM only allows "dial to talk", which is why we need SIP. With SIP, we can do all the above things. However, this is a little too abstract possibly. What do we mean by "click to connect"?

Well, with the Web, we might think of HTTP as the fundamental enabling technology. In fact, this isn't really the case. The essence of The Web is the "universal client" paradigm, or the browser. With a browser, we can access all kinds of data from all kinds of sources (courtesy of HTTP of course, and HTML).

What is the equivalent of the browser in the SIP world? Well, there isn't one really, which is why I have pointed out before the problems of the SIP-HTTP analogy. I restate the problem here:

The Web = HTTP + Browser + HTML

SIP = SIP!

Substituting IMS doesn't work either:

IMS = SIP + Some interfaces (e.g. Billing)

What's missing? Well, the crucial component is the user interface.

Language escapes us at this point, because there is no word to describe the forthcoming SIP-based user experience. However, the missing ingredient is something called Presence. If I stick with the view that I started with, then I believe that the future of the 3G era is:

3G = SIP + Presence

or,

3G = IMS + Presence

which is why I said that 3G is IMS, even though a lot of the current obsession about the future is with speed, or something called HSPDA. This means very fast download speeds, but that's not really a paradigm shift: it's not a new engine, just better fuel.

The way to think of this new paradigm is to think of Instant Messaging, specifically the buddy list component. In the buddy list we can see our buddies and their presence condition: online, offline, busy, away, out-to-lunch, etc.

In future, this is how all of our connecting will be driven - via a buddy interface, except that we shall have more than the option to "click to IM". IM will be only one component. We will have all the other options mentioned above, including "click to talk", which will connect as for a voice-call and replace the "dial to talk" mode that we use today.

We could also "push to talk" (like a walkie-talkie), "push to view", "click to play", "click to share" (i.e. contacts, notes etc.) and all manner of other methods of connecting. In fact, the "connecting" paradigm will extend beyond P2P exchange. Machines will begin to appear in our "buddy" icons, such as our home security system. We can imagine a buddy called "Alarm" and we can see its state: "armed", "unarmed", "red alert", "yellow alert" and so on. Another example is a taxi buddy, a flight buddy and so on.

In some cases, the presence state is the information that we require and is a subtle form of communication in itself and one that we shall become increasingly reliant upon as the process of mobilisation marches on. In other cases, the presence information is an aid to connecting thereafter and the two are closely related. Together they still constitute "connecting".

With the personification (reification) of objects (e.g. machines), we might reasonably expect that dialogue-based communications might become a familiar mode of interaction with machines instead of the current transaction-based (e.g. Web-control) modes. For example, we could imagine having a dialogue with our alarm system via IM. It would "speak" to us as if it were human: "The system is currently unarmed, would you like me to arm it?", or "The alarm has been reset, would you like to call your neighbour?" and so on. In many cases, this is a far more efficient mode of communication than messing around with web-based options and all that non-linear navigation that goes with it. However, it is also possible to move from one mode to another.

With all these modes of connection, SIP allows all the underlying connections and signalling to take place, including transfer of presence-state information. Presence, by which I really mean "buddy-centric" communication (people or object), is an essential component of mobile computing, as it really provides the "Universal client" (and metaphor) through which we shall interact with the digital world.

Mobilisation is the name of the process of folding more and more of our daily tasks into the mobile computing realm. This is a two-way process. Technology improves and produces enablers. Circumstances change, economically, socially, psychologically, that lead us to discover how the enablers might be useful to manage aspects of our changing world.

The buddy-driven presence paradigm will play a significant role in the mobilisation process, if only because it provides us with a model of the world ("world view") that we can work with through our mobile computers. Connecting with "buddies" seems a very natural paradigm.

IMS allows operators to build an infrastructure that will support this paradigm. Therefore IMS will become important in the future of the 3G era. 3G is IMS.

 

0 Comments:

Post a Comment

<< Home