for early service trials, and phased deployment, but eventually, users need to be able to migrate across service usage and platforms in real time. Additionally, Security in general and Identity Management in particular have become central requirements. Customers are demanding location and presence as service enablers. Opening the development environment to external developers greatly increases the eventual services available to the provider’s customers and is expected to lower overall costs. However, this is not a wall-less garden – providers still have control of what is deployed. This adds a requirement to account for usage of and bill for services not originated by the provider.

There is also a strong current association between IMS and SDPs. Alan Quayle characterizes the kinship between SDP’s and IMS:

“SDP Approach:

• The SDP aims at optimizing the operator’s IT and service layer infrastructure by replacing a great number of existing “stove pipes” by a single “horizontal” service delivery environment.

“IMS Approach:

• The IMS aims at optimizing the operator’s network infrastructure by providing a single, all-IP core architecture for all types of exciting and future access networks.”

Alan Quayle uses an apt example to explain differences in IMS and SDF, while underscoring their sibling nature. He sees IMS as the province and program of the service provider CTO; while SDP is the solution put forth by the CIO. But the hope is that SDPs and IMS will converge. SOA technology is seen as the enabler of this convergence, but SDPs are such a diverse lot at present that their business goals are frequently the only thing they have in common. SOA is not always used. Indeed, many have little to do directly with IMS or even SIP. Use of Parlay and Jain is commonplace. SDP’s are being built, but every one is different and likely incompatible. Increasingly, this incompatibility is seen as a problem. It limits assembly of end-to-end services which cross platforms and providers. It greatly increases the expenses of deployment and in particular the management costs. Increasingly, service providers want a managed service from SDPs.

These diverse SDPs need to converge. Tony Richardson:

“SDPs etc are being developed but there is no overall framework for interoperability, agreed capability etc. This could well lead to stove pipes of future services. One of the objectives of SDF is to prevent this – and in