At the time, some industry watchers weren't all that impressed with the idea because they thought IMS obviated the need for such a device. But with IMS adoption occurring at a slower pace than expected, analysts and network operators now are starting to give Tekelec's idea more thought.

"In a way, Tekelec's message has been validated by the market," said Joe McGarvey, principal analyst of IP services infrastructure for Current Analysis. "A year and a half ago, their story may have seemed tired or out of touch with the market, but as IMS adoption has been postponed, it has become clear that the transition will be more incremental and more gradual, and carriers are going to have to live longer in a hybrid world."

Tekelec claims it is gaining traction among network operators with an approach that bridges NGN architectures and IMS. The company made its SIP Signaling Router (SSR) generally available early this year and says it now has a handful of customers worldwide using the product, although so far company officials have not been able to announce any of those customer wins.

The SSR was developed from Tekelec's TekCore Call Session Control Function (CSCF), which is part of the company's IMS architecture, said Vince Lesch, vice president of product marketing for Tekelec. "We were involved in some wireless trials with the CSCF, but increasingly we were seeing that carriers were deciding to put off their IMS deployments," he said. "We began talking with them about their current needs and started to see how we could reposition the TekCore CSCF as a signaling router and have it deployed in existing networks."

The problem with current NGN architectures is that they have no core signaling infrastructure, which limits their ability to scale, Lesch said. In an NGN, routing data is provisioned on each network element, and the data must be updated frequently to reflect changes in network topology. That can lead to routing table exhaustion. The SSR does for the NGN what Signaling System 7 (SS7) and signal transfer points did for the voice network, Lesch said. It pulls out that signaling information and centralizes it so that it can be easily updated.

"Tekelec has a unique angle of attack in the IMS marketplace," said Ronald Gruia, principal analyst and program leader for emerging telecom for Frost & Sullivan. "The traditional equipment vendors like Alcatel-Lucent and Nortel are trying to position the central office switch as the optimal point from which the network will evolve, while Tekelec, with its strength in the SS7 space, is claiming that maybe signaling is the better place to start," Gruia said. "By using Tekelec's approach, a service provider can add some new blades to what they already have deployed and then offer new services while still bridging the gap with legacy services."

That ability to provide a stepping-stone approach to IMS is appealing to network operators. Eric Bozich, vice president of product management for Qwest Communications, said it's crucial that network operators be able to get more out of their existing infrastructure investments before moving to IMS.

Q: What are the chief obstacles a company such as Qwest faces in moving from a traditional TDM network to IMS?

A: One of the challenges lies in making voice over IP (VoIP) an alternative that has all of the same availability and reliability characteristics of TDM voice. The interconnects between carriers just don't exist yet for VoIP that have existed for 100 years in the TDM world.

Q: What can you do to bridge the TDM and IP worlds?

A: It's not so much a technical challenge. We know how to connect customers who have VoIP to the [public switched network] through gateways. But the problem is that the VoIP footprint doesn’t line up with the [public switched network] footprint, and the economics just aren't there to support extending the VoIP footprint to high-cost areas.

Q: What does IMS mean for service providers?

A: Before making an upfront investment in anything new, there's got to be a business model that supports the investment. IMS has role to play in stitching all of this stuff together and creating this concept of seamlessness. But there are pretty substantial investments to make to do that, and it's got to be done in an environment where the new IP-based services are expected to be delivered at a lower price. It's pretty hard to get the economics to work out on that.

Q: Do you think the move to IMS is taking longer than expected?

A: IMS definitely has not lived up to the hype from the early days, and it is clearly progressing at a much slower rate than initially expected. But it's one of those things that will happen in due time.

Q: Tekelec is promoting the idea of a kind of signal transfer point (STP) for the SIP world, which could be used to bridge the gap between legacy and IMS networks. You said you do not want to comment specifically on Tekelec's products, but what do you think of this idea?