SIP, the enabler

SIP (Session Initiation Protocol) was actually late to the VoIP party. The original IP PBXes were mostly based on a collection of International Telecommunications Union standards called H.323. SIP, a text-based protocol that draws on HTTP and MIME, grew out of research at Columbia University and didn’t become a real force until it was adopted by the IETF in 1999.

“Both SIP and H.323 do a reasonable job setting up voice and video sessions,” says John Yoakum, director of emerging opportunities in enterprise networks at Nortel. “But SIP is the more expandable and extensible framework that will bring us into the next generation of communications technologies. The war between SIP and H.323 is over, and SIP has won.” In fact, most IP telephony vendors are incorporating SIP into their offerings, often in addition to H.323.

The beauty of SIP is that although it enables typical IP telephony functions, it’s not a VoIP standard but simply a protocol for setting up and managing IP-based communications sessions involving anything from voice, video, and IM to file exchange, games, and whatever else comes along. In fact, the latest version of Microsoft’s LCS (Live Communications Server) is SIP-based, and Microsoft has incorporated SIP into Windows CE, Windows XP, Windows Server 2003, and Windows Messenger. The IETF also has some extensions to SIP in the works specifically for IM called SIMPLE (SIP for IM and Presence Leveraging Extensions).

SIP is device-agnostic, allowing SIP-enabled notebooks, cell phones, IP phones, PDAs, and set-top boxes to negotiate communications directly with one another based on common capabilities. Similar to DNS, SIP can use a dynamically updated registrar to map devices to names so that you can address communications to a person, regardless of the device or network that person happens to be using at a particular moment. SIP’s inherent flexibility and extensibility makes it a natural for IP-based communications and applications.