A chat earlier this summer with Cisco Fellow Jonathan Rosenberg helped plot a few trends lines for session initiation protocol (SIP).
Also commonly known as request for comment (RFC) 3261 of the Internet Engineering Task Force (IEFT), but comprised of some 150 total RFCs, SIP is gaining traction within business communications.
“We’re definitely seeing a movement toward SIP within the enterprise, all around,” said Rosenberg.
Before moving into Cisco’s enterprise voice group two years ago, Rosenberg said he was seeing SIP-based interconnects as a cost-reduction move among service providers, including MSOs. Within the business arena, however, SIP is moving closer to the end-user.
That arena has grown busier of the past few years. “Increasingly, communications is not just the PBX, and that’s it,” Rosenberg said. “You’ve got VoIP, IMS, presence, separate voice mail and conference meeting servers.”
The IT challenge for enterprises is interoperability, or plugging in endpoints from multiple vendors. “SIP has been the glue for some time,” Rosenberg said.
Rosenberg pointed to Avaya as a company that has pushed next-gen SIP for the enterprise. At VoiceCon 2009, the company announced a “multi-vendor, multi-location, multi-modal” SIP-based architecture called Aura.
Another area of SIP enterprise growth is comparable to the interconnect applications among service providers. “SIP trunking is still early, but it’s on an uptick especially in the larger enterprise, (for instance) into centralized data servers,” he said.
(For more on SIP trunking for cable, click here.)
Within a network context, session border controllers (SBCs) are part of the reality of getting SIP in practice, Rosenberg also noted.
In a related research note posted last week at No Jitter, Zeus Kerravala, a Yankee Group VoIP analyst, expanded on this connection between SIP trunking and SBCs. “The biggest area of ‘low hanging fruit’ for this (SBC) market is in helping enterprises connect to SIP trunking providers.” (For more, click here.)
In light of SIP’s role as glue within an IMS framework, Rosenberg listed two additional drivers, both in the wireless space. First: Long-Term Evolution (LTE). Once LTE deployments actually happen, they “become a point of no return…(and) a driver for the core IMS.”
“Another is applying IMS to existing cellular circuit voice endpoints among the wireless carriers,” he said. “Take existing cell phones, put IMS on top of that, as a sort of transition-type technology.”
The specification in play for that application is IMS centralized services (ICS). For more comments on SIP and IMS, some from a Comcast perspective, click here.)