Transforming telecoms into next-generation networks with SIP may sound like a gamble, but Sonus Networks is a company
that likes to make big long-term bets, according to Vikram Saksena, the company's CTO.
"We try to be a leader in the space we go after," he said. "That's what we did 10 years ago with VoIP, and now we're trying to make another long-term bet." This time, Saksena and Sonus are hoping that emerging services based on Session Initiation Protocol (SIP) and related protocols take off and help telecoms better manage – and monetize – the user experience.
Currently, Sonus sells technology that provides both landline and wireless VoIP capabilities to carriers like AT&T and BT. Saksena said the company is taking what it learned with VoIP management and applying it to a wider variety of multimedia content, while investigating ways to open up its framework to third-party developers.
It's not a particularly new story, but as carriers look to make their own services stand out above Internet-based competitors, any chance at jump-starting their networks is worth a look. Some worry that the evolution will not occur fast enough or in the right directions for companies like Sonus.
"The trouble with SIP is that it is a connection signaling protocol. It's not a service signaling service program," said Tom Nolle, president of telecom analysis firm CIMI Corp. "The general view I have is that if you're going to give somebody Internet access, session-based services border on non-merchandisable."
Users have become too used to open-ended, all-you-can-eat connections instead of the discrete services that telecoms are used to building and selling, and to get customers to revert to the old ways is a tough sell, which is where Saksena's role as a visionary comes in.
Many of the advanced features that will set telecoms apart will be focused on the embedded intelligence of next-generation networks, Saksena said. Think targeted advertising, or allowing viewers to see which shows are popular with their friends or neighbors.
"There used to be a context of intelligent network in the old days of PTSN, but those never did anything very intelligent beyond caller ID and call forwarding," he said. With the next generation, however, intelligence will mean having a profile of what kinds of content and services the user likes, where the user is located, as well as the ability to link experiences across Web browsers, mobile phones and HDTV.
"If you want to own the entire consumer experience at home, then you have to be a converged operator," Saksena said. This is a critical part of his vision for the future: no wireless or landline providers; just service providers that let users connect in a variety of ways.
Saksena is also trying to address the groundswell phenomenon of Internet generativity. While telecoms have always been slow to develop and deploy new services and products, nimble Web 2.0 companies have surpassed them and cut into the profitable niches like video and voice that telecoms once had locked down. Developing competitive products entirely in-house is inefficient and ultimately ineffective for telecoms, Saksena said. They will find more success by providing a framework to outside developers that includes a monetization strategy to cut providers in on a slice of the profits.
"The telecom guys are looking at the Internet model of service creation," he said, noting that BT was experimenting with developer communities. "That's where the long-tail thrives: They build applications over time and they catch interest."
None of these shifts will require sudden overhauls in Sonus strategy or architecture, Saksena believes.
"I think this is how we augment our products; these are some additional capabilities we need," he said. "So far, we don't need anything radically different; just evolve the products with the right architectural thoughts."
It is a strategy that the market is finally ready to follow, according to Jeff Ogle, principal analyst with Current Analysis.
"Telecoms are coming to that reality," Ogle said. "They've been looking at it for the past several years. [AT&T's multimedia offering] U-Verse … was in place for years, when carriers were trialing things in specific regions."
The early leaders to help make the transition will be companies like Sonus and ShoreTel that have a strong IP background, Ogle said.
"A lot of the smaller guys are pushing this; a lot of people who came from an IP background are pushing this," he said, "They're not your traditional wireline companies. They're not your Nortels or even your Ciscos, though Cisco's jumping on the bandwagon."
Nolle said, however, that even these few years of legacy IP technology may be holding back some companies if they are too invested in their own technology.
"The IMS people and SIP people are going to have to reorganize their thinking to the open handset movement and convergence," he said, adding that these concepts will have profound impacts that he thinks most companies still underestimate. "I just don't think they're going to change their ways fast enough to be survivable."
Part of the problem is that many of the features Saksena touts – interoperability, open APIs, service control – are already available in many other standards, and SIP often is not the best way to go about implementing them.
"There's no lack of capability in that area. You don't have to do it with SIP," Nolle said, pointing to the Parlay vendor consortium and its XML-based APIs, which are already embedded in many networks and don't require major new infrastructure investments. "The real challenge is to not expose the customer, network or third party using them to unreasonable risk."
Nolle said that by going forward and updating their legacy technology, these companies will probably end up dooming themselves by tying their fortunes to specific architectures that may not be in demand in a few years.
"Smaller players are going to pitch in and do something here," he said, "because the larger, more established players just can't turn in the right direction."