The skinny on SIP

This week, we’d like to discuss Session Initiation Protocol, a critical standard for moving voice over IP to the next level.

While we’re certainly pleased with the industry’s adoption of H.323, and while we believe both protocols will continue to co-exist as underlying standards for VoIP, broader deployment of SIP will enable an even greater range of integrated communications features and services.

SIP Version 1 was first introduced in 1996 to the IETF as Session Description Protocol (SDP), relying on User Datagram Protocol (UDP) as a transport mechanism.

The IETF combined SIP Version 1 with a similar protocol called Simple Conference Invitation Protocol (SCIP) to create SIP Version 2. By adding SCIP, SIP was now able to use HTTP and Transmission Control Protocol (TCP) as transportation mechanisms.

As a signaling protocol, SIP provides the network with the capability to establish, modify, and terminate multimedia sessions between user applications like voice and audio, video, shared data applications, and gaming.

SIP uses addresses similar to e-mail addresses to identify users. These are listed in the format of SIP URLs. For example, Larry’s SIP URL might be SIP:Larry.Hettick@larryhettick.com. When Larry logs on to a given domain, his presence is registered by the network. When Steve places a voice call to Larry, the network finds Larry’s domain included in Larry’s SIP address, notes his registered presence at the given domain and establishes the call.

But what happens if Larry is logged onto more than one device or is present in multiple domains? For example, Larry can be reached on his IP phone and on his wireless phone since he has both devices registered at the same time. What happens if Larry changes locations? Larry will most certainly not sit still at his desk all day. Fortunately, one important aspect of SIP is the ability to support user mobility.

Next time, we’ll talk about how SIP can direct and redirect calls to the right place and the right device.