Next: 3.1.3 Architecture Overview Up: 3.1 Videoconferencing and VoIP Previous: 3.1.1 Overview
The ITU Telecommunication Standardization Sector (ITU-T) is one of three Sectors of the International Telecommunication Union (ITU) and works to provide high quality standards (Recommendations) covering all fields of telecommunications. Arguably the most popular and extensible early compressed videoconferencing was enabled via the ITU standard called H.320, describing videoconferencing services over ISDN and T1 leased and dedicated telephone lines.
H.323, was first approved by the ITU in 1996. H.323 was designed to both function like and interoperate with H.320, changing the transport layer only so that the protocol would work on the Internet. H.323 parallels the telephone system in having an architecture based on intelligent call control and relatively "dumb" endpoints; this design is based on circuit-switching and a desire to provide centrally controlled, highly managed services.
H.323 is not one particular protocol but is rather an umbrella standard consisting of several different protocols. Signaling and call control are handled by Q.931 as defined in Recommendation H.225.0 and Recommendation H.245. Once call control completes, the media transfer can begin by using the IETF Real Time Protocol (RTP). A variety of audio and video codecs are supported, and the standard includes both necessary and optional components. H.323 includes a security protocol known as H.235, which describes authentication and encryption using passwords or digital certificates.
SIP is a signaling protocol for establishing calls and conferences over IP networks. Unlike H.323, SIP is a standalone protocol; it does not handle media transfer, resource reservation, or even session description. Even though SIP is not an all-in-one solution, it does work well with other protocols and thus, allows for flexibility in defining a videoconferencing solution. RTP may be used for media transfer and SDP for session description. SIP originated in the mid 90's (about the time H.323 was becoming finalized as a standard) so that it would be easy to invite people to view an IP multicast session like a shuttle launch on the M-Bone.
Rather than being a telephony-based protocol, SIP is modeled after other Internet text-based protocols such as email (SMTP) and the web (HTTP), and was designed to establish, change, and tear down calls between one or more users in an IP network in a manner totally independent of the media content of the call. Like HTTP or email, SIP moves application control to the endpoint, eliminating the need for intelligence in the network core. This design is based on TCP/IP and a desire to put as much management control as possible at the endpoint.
The SIP protocol includes authentication using passwords or digitally signed (S/MIME) messages. The SIP protocol is fully described by IETF RFC 3261.