How SIP works: Session Initiation Protocol explained

Who this guide is for

Audience: IT administrators and technical buyers learning SIP trunking basics.

Understand how VoIP calling works — SIP, PBX, codecs, trunking, DID numbers, STIR/SHAKEN, and the protocols behind business phone systems.

Use this guide when you want the setup to be correct the first time and easy for another admin, manager, or supervisor to verify later.

What this workflow helps you accomplish

SIP (Session Initiation Protocol) is the signalling protocol that sets up, manages, and tears down real-time communication sessions — voice calls, video calls, and messaging — over IP networks. SIP handles the negotiation between endpoints so media can flow after the session is established.

This workflow matters because numbers, routing, access, and reporting in CallOrbit are connected. Skipping one setup detail usually creates avoidable support work later.

• Step 1: Learn the SIP message flow: an INVITE request initiates a session, the remote endpoint responds with 100 TRYING, 180 RINGING, and finally 200 OK to accept. ACK confirms receipt, and BYE terminates the session. This handshake happens in milliseconds.
• Step 2: Understand SIP addressing: each SIP endpoint is identified by a SIP URI (sip:user@domain). SIP addresses can resolve to desk phones, softphones, PBXs, or SIP trunk providers, making routing flexible regardless of physical location.
• Step 3: Distinguish SIP signalling from media transport: SIP negotiates the session parameters (codecs, ports, IP addresses) but does not carry the actual voice audio. Once SIP establishes the session, RTP takes over to transport the media stream between endpoints.
• Step 4: Recognise common SIP response codes: 200 OK (success), 401 Unauthorised (needs authentication), 404 Not Found (user not found), 480 Temporarily Unavailable (endpoint offline), and 603 Decline (call rejected by the recipient).
• Step 5: Configure SIP registration so endpoints authenticate against a SIP registrar. Unauthenticated SIP endpoints cannot place or receive calls through the provider's network, which prevents toll fraud and unauthorised usage.

Setup checklist

• Learn the SIP message flow: an INVITE request initiates a session, the remote endpoint responds with 100 TRYING, 180 RINGING, and finally 200 OK to accept. ACK confirms receipt, and BYE terminates the session. This handshake happens in milliseconds.
• Understand SIP addressing: each SIP endpoint is identified by a SIP URI (sip:user@domain). SIP addresses can resolve to desk phones, softphones, PBXs, or SIP trunk providers, making routing flexible regardless of physical location.
• Distinguish SIP signalling from media transport: SIP negotiates the session parameters (codecs, ports, IP addresses) but does not carry the actual voice audio. Once SIP establishes the session, RTP takes over to transport the media stream between endpoints.
• Recognise common SIP response codes: 200 OK (success), 401 Unauthorised (needs authentication), 404 Not Found (user not found), 480 Temporarily Unavailable (endpoint offline), and 603 Decline (call rejected by the recipient).
• Configure SIP registration so endpoints authenticate against a SIP registrar. Unauthenticated SIP endpoints cannot place or receive calls through the provider's network, which prevents toll fraud and unauthorised usage.

Operational follow-up

After you complete this flow, confirm the live experience from both the agent and customer side so ownership, routing, permissions, and reporting all match what the workspace expects.

If your team is rolling this out across multiple users, queues, or phone numbers, pair this article with the broader knowledge base and the relevant routing or numbers guides to keep deployment consistent.

• What is the CallOrbit Knowledge Base for? — It is the public help hub for how CallOrbit works, covering numbers, webphone setup, SIP, extensions, routing, users, roles, and billing basics.
• Can customers read this without signing in? — Yes. The Knowledge Base now lives on a public route so customers can read setup guidance before or after they enter the portal.
• Does the portal still have its own Knowledge Base page? — No. The signed-in portal navigation no longer carries a separate Knowledge Base page, and the old portal path now redirects to this public version.
• What is VoIP and how does it work? — VoIP (Voice over Internet Protocol) converts analogue voice signals into digital packets and transmits them over IP networks. Unlike traditional PSTN phone lines that require dedicated copper wiring per line, VoIP calls use your existing internet connection, which makes them cheaper, more flexible, and easier to scale.
• What is SIP trunking? — SIP trunking is a virtual connection that replaces traditional analogue phone lines or PRI circuits. A SIP trunk carries multiple concurrent voice channels over a single IP connection to your PBX or phone system, eliminating per-line hardware costs and monthly line rental fees.
• What is the difference between hosted PBX and cloud PBX? — Hosted PBX runs on dedicated virtual infrastructure managed by a provider, while cloud PBX uses shared multi-tenant cloud infrastructure. Hosted PBX suits organisations needing custom configuration and predictable pricing. Cloud PBX is better for instant scalability and per-user monthly billing.
• What is a DID number? — A DID (Direct Inward Dialling) number is a virtual phone number that routes directly to a specific extension, IVR menu, queue, or user within a phone system without an operator. DIDs decouple the phone number from the physical phone line, so you can have hundreds of numbers routed through a single SIP trunk.
• What are G.711, Opus, and G.729 codecs used for? — These are VoIP codecs that convert voice into digital data. G.711 uses 64 Kbps for toll-grade quality and is the PSTN standard. Opus uses 6-510 Kbps and adjusts to network conditions. G.729 uses 8 Kbps for bandwidth-constrained links. The right codec depends on your available bandwidth and call quality requirements.

Related CallOrbit guides