IMS & The Stupid Network

By Martyn Davies

IP Multimedia Subsystem is championed as the way that telcos can quickly and cheaply bring new services to customers. There’s much to commend the IMS vision, bringing Internet technologies and telephony together. However, much has changed since telcos rolled out the last big thing, the digital switching network. David Isenberg coined the term stupid network to describe a network where the intelligence is in the endpoints, leaving the network itself to be a dumb bitpipe. The Internet sees many new services being launched every week, so we are now seeing the promise of the stupid network, and how it can challenge aspects of IMS, but also invigorate it. A key IMS goal is to offer the same services to customers regardless of the way that they accessed the network. Because IMS came out of the 3GPP committee, initial focus was on access by cellular phones. However, as time went on, telcos could see the applicability to fixed networks, and so this came under the same umbrella. If you talk to British Telecom today they will tell you about their 21st century network, 21CN for short, which is an ambitious plan to tear apart the digital network core and replace it completely with IP by 2010. The success of technologies like WiFi ( News - Alert) and broadband has resulted in these systems joining the IMS umbrella. Mobile computing is a strong driving force, with PDAs evolving phone functions, and at the same time phones becoming smartphones with added e-mail and Web browsing, and with WiFi and Bluetooth. Devices like Blackberry and PocketPC are no longer for geeks, but for serious business use.

So with users logging on whichever way, this should result in a wider user-base for services, and so more revenue, and a faster return on investment. With current mobile phone networks, the creation of new services has been a headache for service providers, and many talk about a new service costing $1 million and service rollout taking one year. Part of this is due to the walled garden nature of today’s phone services, where largely the services you can have as an operator depend on what the core equipment vendors can provide.

Customization can be a headache. Another design goal of IMS is to allow the edge of the network to be opened so that services can be offered not just by the operator themselves, but also by service partners. In other words, this would be more like the highly successful i-Mode architecture in Japan, where the telcos provide an infrastructure, and access to customers, but the applications themselves are provided by separate businesses, with their own business goals, marketing and brand. This way, the telco and the application partner get to share the revenue, giving both an incentive to evolve more services in partnership.

One consortium hoping to build this application infrastructure is the Parlay Group, which is a vendor-independent group, creating standard APIs for applications that run on IMS networks. Parlay includes vendors in the software and services business, for example IBM (News - Alert) and BEA; networking hardware companies, like Ericsson (News - Alert) and Lucent; and also of course telcos like BT and Sprint (News - Alert) . The standardization of a programming API allows for application portability, so that service providers that run successful services in conjunction with, say, Vodafone, would also then be able to market the same service with BT and Sprint, and roll out with a minimum of changes.

One particularly interesting API from Parlay is known as Parlay-X. Zygmunt Lozinski, President of the Parlay Group, described it like this: “What parlay-X is, is a set of Web services, so somebody that’s used to writing Web services, because they’re used to programming with .NET or one of the standard frameworks like that, can build applications that can make use of the enablers within a telco network.”

Web services is a standard way of interacting with application servers using the Web as a transport interface, and specifying the API operations themselves in the form of XML. It is the basis of many of the so-called Web 2.0 APIs, such as those published by Google 
( News - Alert) to their developer community. Large software companies using this approach to deliver large-scale IT applications include Oracle (News - Alert), IBM, Microsoft (News - Alert), and BEA.


Lozinski again: “[In the IMS network] you have a set of application servers, that can reside in the network, or they can reside outside of the network, that’s a business decision by the operator.”
In other words the applications that the end-user sees can be from entrepreneurial, independent companies that are using the telco services as a component.

But enough of IMS, what of the stupid network? An obvious starting point when talking about the stupid network is the Internet softphone, Skype (News - Alert). Skype is a free download, and after you register a unique name for yourself, you can immediately start calling other Skype users, free of charge. Skype needs no intelligent network core, only a bitpipe with enough bandwidth, so it can be used over broadband, over WiFi, even over business VPNs. The customers do not need special telco services with extra monthly fees, they just log on. This is the power of the stupid network, that by running different software at the edge of the network you can have any capabilities you like, as long as your friends and colleagues have the same software.

Shel Holtz and Neville Hobson are prolific PR Bloggers, and also run a bi-weekly podcast, For Immediate Release: The Hobson & Holtz Report, about the world of business, technology, and PR. Enthusiastic users of stupid network services, they use Skype in order to make the show. With Holtz in Concord, CA and Hobson in Amsterdam, they need to connect up in order to record the show. As Holtz says: “We are on Skype something like six hours a week, sometimes more, and if we had to pay the cost of a transatlantic call from the San Francisco Bay Area to Holland, that would add up to some considerable charges.” If they had to connect up using a traditional transatlantic call, he doubts whether their podcast would have been born. But another factor is actually quality “If it sounded like I was in my office and Neville was on the phone, I don’t how many people would listen to that for 90 minutes twice a week.”

This highlights another advantage of the edge approach: services like Skype are free to use wideband codecs to improve the voice quality, and not limited to traditional telco bandwidth. Furthermore, if Skype improves or adds more codecs, users can access these In many ways you can look at today’s stupid network services and imagine that tomorrow’s IMS services will have the same characteristics.straight away simply by downloading a new client; they don’t have to ask a telco for permission, or spend more money on it.

In many ways you can look at today’s stupid network services and imagine that tomorrow’s IMS services will have the same characteristics: They should innovate, evolve quickly, and it should be possible to try and reject ideas in a way that does not break the bank. Stupid network services are often more cleverly marketed than their telco-based alternatives. Services like ICQ (now AOL Instant Messenger) and Google’s G-mail became successful due to word-of-mouth, which quickly spread the message to the potential audience.

Holtz again: “I think Skype has been exclusively word-of-mouth. I’m not aware of any other advertising or marketing they have done, compared to Vonage (News - Alert), whose acquisition and marketing costs are huge, and haven’t attracted anywhere near the user base that Skype have.”

In the Internet world, the business model that IMS strives for is already well advanced: as Holtz says: “In this world, dare I say it, of Web 2.0, there are companies releasing the APIs to their application: for example Googlemaps; the folks at Frappr have taken that API and mashed it up, and made it something that’s useable.” And we can already see the genesis of this approach in the IMS world.
Lozinski again “The [IMS] core network has scalability, robustness and accessibility to everyone, and this can combine with the interaction possibilities of the Internet.”

Lozinski describes a Parlay-X application, which involves placing yellow arrows, each marked with a code and SMS number. By sending the code by text message, you can find out what information has been associated with the arrow, and that geographical spot. For example, you might attach these arrows to restaurants, and this would allow people to get the last five reviews that visitors had submitted for that restaurant. “That’s combining the fact that telcos have SMS and that sort of position information with Web 2.0-style community interaction, where all of the data around that point is of interest.”

The IMS network provides unique value in terms of robust services, and information that cannot be got in other ways, for example location and user availability information from the mobile network. On the other hand, the power of services like Skype and Google is in the flexibility, coupled with efficient marketing techniques, and a reach of millions of customers. By exposing APIs to the outside world, and working with Internet partners outside the walled garden, IMS can enable the kind of Web 2.0-style co-operation, bringing the Internet buzz and telco customers together. Perhaps the IMS and stupid network approaches are not so far apart after all.

Martyn Davies is a principal consultant for Eicon Networks (News - Alert) . For more information, please visit www.eicon.com.

[Return To The Table Of Contents ]