Hosted VoIP

 

In the late 1980s and through the 1990s the telecom industry defined and then adopted standards for the “Intelligent Network” (IN). The objective was to make it easy to create and deploy new applications and new services. More recently the industry has defined various next generation networks, most notably the IP Multimedia System (IMS). Besides convergence economies (all services on IP), the objective is to facilitate the rapid development and deployment of new applications and new services. Sound familiar? Unfortunately, it’s not going to work any better the second time around.

 

Mind you, the Intelligent Network was not a failure. While it didn’t foster a zillion, or even a hundred, new applications it did support massive scaling of few applications, for example, “Freephone” service (800 numbers), roaming (the basis for mobile telephony), voicemail and prepaid.

 

The first problem is business models. Intelligent Network functionality is service-affecting and therefore carefully controlled. Outside of a few large equipment suppliers, developers don’t get access to the IN. As a result, most new services are invented at the edge. For example, voicemail originated as an enterprise application and was only integrated into the network many years later. Even prepaid phone service was implemented from the edge and only integrated into the IN over a period of years.

 

IMS is following exactly the same path. Again, the network-based SIP application servers that define IMS call flows are critical to basic telephony. As a result, no operator lets third-party developers run applications on their IMS core. Even worse, most operators maintain a walled garden for data access. This means you can’t get IP connectivity without a special business relationship. At least with the IN you could get access at the edge (via PSTN phone numbers). With IMS, you can’t even count on access to the edge of the operator’s IP network.

 

The second problem is service creation software. Unfortunately, for both the IN and for IMS, the standards focus on basic call control and media flows; i.e., on the plumbing. There are no IMS standards for software APIs or software development environments. Typically, each major equipment provider has a set of service creation tools, but they differ from vendor to vendor.

 

So what’s the most likely outcome? Like the intelligent network, IMS will be used to deploy and massively scale a few applications over the next decade or so, but for innovation, we will have to look to the open Internet where anyone and everyone gets to experiment with their idea of a great new service.

 

-----