Session Initiation Protocol

Session Initiation Protocol
source
Name	Session Initiation Protocol
Developer	IETF
Introduced	0 1999
Osi layer	Application layer
Ports	5060, 5061 (TLS)
Rfcs	RFC 3261, RFC 6665, RFC 3262, RFC 3263, RFC 3264

Session Initiation Protocol. It is a signaling protocol standardized by the Internet Engineering Task Force and is fundamental to modern Voice over IP communications. The protocol is used for initiating, managing, and terminating real-time sessions that can include voice, video, messaging, and other communications applications over IP networks. Its design is text-based, similar to the Hypertext Transfer Protocol, and it operates independently of the underlying transport layer, supporting UDP, TCP, and TLS.

Overview

Developed within the IETF's MMUSIC working group, the core specification is defined in RFC 3261. It was designed to be a modular component within the larger Internet Protocol Suite, specifically for creating and controlling multimedia sessions. The protocol's architecture is inspired by elements of Hypertext Transfer Protocol and Simple Mail Transfer Protocol, utilizing a request-response model and Uniform Resource Identifiers for addressing. Its primary role is in the establishment of sessions, while other protocols like the Real-time Transport Protocol typically handle the actual media stream. This separation allows for flexibility and integration with various services and infrastructures, such as those defined by the 3rd Generation Partnership Project for IP Multimedia Subsystem networks.

Protocol operation

A typical operation involves a series of transactions between logical entities, beginning with an INVITE request from a user agent client. This request is routed through the network, potentially interacting with proxy servers and registrar servers, to reach the intended user agent server. The protocol supports mechanisms for locating users, such as those provided by Domain Name System servers and ENUM lookups. Key functions during session establishment include negotiation of session parameters using the Session Description Protocol, which is carried within the protocol's message body. For reliability, it incorporates mechanisms like provisional response acknowledgments, detailed in RFC 3262.

SIP messages

Messages are categorized into requests and responses, following a format that includes a start-line, headers, and an optional message body. Fundamental request methods, defined in the core RFC 3261, include INVITE, ACK, BYE, CANCEL, OPTIONS, and REGISTER. Subsequent extensions introduced methods like MESSAGE for instant messaging, PUBLISH for event state, and SUBSCRIBE/NOTIFY for event notification, as standardized in RFC 6665 and related documents. Responses are three-digit numeric codes, grouped into provisional (1xx) and final (2xx through 6xx) categories, indicating success, redirection, or various client and server errors. The message body often contains a Session Description Protocol payload for media negotiation.

SIP network elements

A network based on this protocol comprises several key logical components. A user agent is an endpoint that initiates or receives sessions, such as an IP phone or a softphone application. Network servers include proxy servers, which route requests and enforce policy, and registrar servers, which process REGISTER requests to bind a user's address to their current location. A redirect server provides alternative address information without forwarding requests. For managing complex call routing and features, a Back-to-back user agent can act as an intermediary. These elements often work in concert with other systems like a Location server or a gatekeeper in H.323 networks.

Security considerations

The protocol includes a framework for authentication, integrity, and confidentiality, leveraging existing Internet security protocols. Digest access authentication, based on a challenge-response model, is commonly used. For encryption of signaling, Transport Layer Security can be employed on the transport layer, while Secure Real-time Transport Protocol is recommended for securing media streams. Threats such as spoofing, replay attacks, and denial-of-service are addressed in the core specification and further detailed in documents like RFC 3261. Secure practices often involve the use of VPNs and firewalls configured to handle the protocol's dynamic port usage for media.

Applications and deployment

It is the foundational signaling protocol for most modern Voice over IP and Unified communications systems, including commercial services from providers like Vonage and Skype for Business. It is a core component of the IP Multimedia Subsystem architecture promoted by the 3rd Generation Partnership Project and European Telecommunications Standards Institute for next-generation networks. Beyond telephony, it enables Video conferencing systems, Instant messaging platforms, Presence information services, and integration with emergency calling systems like E911. The protocol is also widely implemented in open-source projects, such as Asterisk, and is integral to WebRTC frameworks for browser-based real-time communication. Category:Internet protocols Category:Voice over IP Category:Application layer protocols Category:Telecommunications standards