By Ken Rehbehn, research director, Current Analysis
When they first came out, one of the coolest things about cell phones was the almost magical ability to reach out to somebody (anybody, really) and talk--regardless of where you were. It was cool because you were on the move. Yet, more than just being cool, it made a difference to communications because it was a direct personal connection with another human…and it worked because both parties had devices and services capable of making a connection over a network. In the parlance of Metcalfe's law, the value of the network was high.
Fast-forward 20 years or so and those old PSTN/cellular circuit-switched networks seem oh-so-quaint. Now, high performance multimedia services delivered over mobile and fixed IP networks promise the next level of person-to-person communication. Video sharing and gaming stand out as two possible applications benefiting from the new era of digital communications. Making such peer-to-peer services--as well as client/server applications--work as part of a service offering is the job of IMS, with extensive building blocks for session control, AAA, security and legacy interconnection. Voice services offered over IMS architectures are already starting to make a presence. Video share services are as well.
However, all is not well in the world of IMS. Despite early promises of IMS serving as a conduit for a wide-ranging set of person-to-person interactions, the client implementations have been limited. In part, this is due to the early stage of the game. But it's also do to complexity. A simple "plain old telephone" and even its complex cousin, the cell phone, enjoy a trivial user interface with sparse--but universal--functionality. If you know a phone number, you can call it. Again, the network value was high.
IMS clients are not simple. Successful completion of a person-to-person interaction with rich sound and graphics over an IMS-based service demands an extraordinary degree of luck. Take an IMS-based chess application. Each person must possess a device (a handset or fixed terminal) that is compatible with the device of the other player--there is little point in thinking about video share if your partner's device lacks a video camera. Often, this goes beyond having the right device to having the right network (supports IMS, supports interconnection to other IMS networks, etc.), the right tariff ("yes, you can play chess for $1.00"), and the right application (both sides knowing how to signal a piece moving). You have to guess whether a friend has the "right stuff" before attempting a friendly game of chess, but with so many variables it is likely you will guess wrong. The result is a user community--everyone sharing the "right stuff"--that is incredibly small. In terms of Metcalfe's law, the value of the network is low. Small wonder we have not seen an explosion of jaw-dropping apps.
So, does this mean that IMS will fail to deliver on high network value for rich person-to-person communication? No.
IMS remains a compelling solution for managing session-based connections over an IP network in the presence of stringent requirements for AAA, security, monetization and legacy interconnection. Communities of interest supported by the "right stuff" in the form of compatible networks, devices and application clients reap the benefit of reliable person-to-person communication. Video share is a case in point. Over the past two years, the GSM Association expended significant energy into making sure W-CDMA mobile devices delivered on the promise of IMS-based video share. Months of testing by numerous network operators, infrastructure vendors and handset suppliers systematically attacked the problem. At the 3GSM World Congress in Barcelona, the GSMA and its members demonstrated the fruits of the effort, raising the prospect of a large potential video share community, one that satisfies the conditions for a high value network.
Just as encouraging, many interesting IMS applications do not depend on communication between two subscribers outfitted with unpredictable devices and applications. Services offered on application servers--leveraging IMS enablers such as presence and location--depend only on compatibility with a subscriber's device and application client. While a person may not know which friends share a common feature, IMS clients/servers know exactly how to find each other. A subscriber with a weather client on a device predictably finds the desired forecast from an IMS-hosted server. Likewise, new approaches towards IPTV also promise to deliver on this concept. Feature interaction, so-called "caller ID" combined with a televised multimedia stream, delivers on the vision of IMS to a well-defined community of users.
Nonetheless, pure person-to-person interaction built on an IMS framework is likely to remain limited to a narrow set of exhaustively proven, and relatively simple, applications. Perhaps incorporation of seamless "just-in-time" IMS client downloads, much as with the fluid web-client paradigms of Web 2.0, can deliver a ubiquitous peer-to-peer application experience that delivers on the power of Metcalfe's law. Lacking such innovations, subscribers of the future are more likely to take advantage of today's well-understood peer-to-peer application mechanism ubiquitous today: the Web browser supporting java script.
Ken Rehbehn is a research director at Current Analysis, a Washington, DC-based independent research firm.