Call forwarding on busy line in ISUP
When a called Australian number is already engaged, call forwarding on busy line (CFB) lets the network divert the call to another endpoint instead of returning a busy indication. In ISUP, exchanges recognise the busy state, locate the forwarding destination, establish a fresh leg and connect the parties. The caller normally hears the forwarded destination ringing rather than a busy tone.
The procedure is more than a lookup that changes a destination number. A forwarded leg carries its own routing, charging, screening and release requirements, while exchanges must prevent loops and inappropriate connections. The flow below follows typical ISUP behaviour and notes where carrier implementations can differ.
Busy-line forwarding in Australian networks
Australian organisations in Sydney, Melbourne, Brisbane and Perth often use PBXs, contact-centre queues and multi-site reception numbers where an occupied extension creates genuine busy conditions. For a company using a central receptionist, CFB can send overflow calls to a mobile, another site or an attendant console without exposing internal extension numbers.
Carrier numbering also shapes the feature. The +61 country code and Australian geographic area codes must remain correct after a redirect, and number portability can move a service between providers. Telstra, Optus and Vodafone may apply different subscriber and inter-network rules, so testing must cover the actual carrier and destination type.
Administrators should check whether CFB is permitted for 1300 and 1800 numbers, premium services or channels with complex IVR menus. A redirect that works on a standard fixed line may fail when operator screening, charging rules or customer premises equipment impose different conditions.
ISUP signalling sequence
A typical flow starts with an Initial Address Message (IAM) from the originating exchange to the terminating exchange. The IAM carries the called number, calling-party identification and other setup data. At the terminating end, the exchange finds that the subscriber line or extension is busy and checks whether CFB is active.
When the service is active, the terminating exchange can send a Call Proceed (CPG) message carrying a busy cause and redirect information to the originating exchange. This identifies the forwarding number and may include presentation or redirect-reason data. Some carrier switches perform the redirect internally, but the logical result is a new call leg rather than the original busy call simply being answered.
The exchange creating the new leg performs routing and screening before sending another IAM to the forwarded destination. If it answers, Answer Charge (ACM) and Answer Message (ANM) processing marks the forwarded call as accepted and connected. Release (REL) and Release Complete (RLC) messages clean up the state. A failure or timeout must release both legs, preventing a phantom connection.
CFB compared with other forwarding services
The key distinction is the event that triggers the redirect, not merely the destination number.
| Forwarding service | Trigger | Typical treatment | Principal concern |
|---|---|---|---|
| Unconditional | Arrival of the IAM | Redirect without waiting for user availability | Accidental diversion |
| Busy line | Confirmed busy condition | Redirect using the busy cause and forwarding data | Loops and duplicate charging |
| No answer | Timed lack of response | Redirect after the relevant timer expires | Delayed calls or voicemail conflicts |
CFB differs from unconditional forwarding because the called party is first found busy; it differs from no-answer forwarding because silence does not invoke it. A swap is not CFB either: it redirects before the user answers and follows different call-handling rules.
Voice diversion can create loops, such as A forwarding to B while B forwards back to A. Redirect counters, loop detection and subscriber screening must therefore be verified, especially when both a PBX and a carrier feature are active.
Routing, transport and security controls
The redirect is reliable only when signalling can reach the forwarded exchange. MTP3, SCCP and route selection must direct the second IAM consistently with the voice path. A design may separate narrow TDM or IP signalling backhaul from switched traffic, so route planning must account for latency, congestion and failover.
At service-logic level, database queries can use TCAP while ISUP manages the call. A worked TCAP component sequence illustrates the contrast between the transaction layer used for subscriber or service controls and the message flow that establishes the forwarded speech path.
Security controls must match the redirect decision. Exchanges should validate forwarding numbers, restrict presentation data, screen premium and overseas destinations, and keep records for disputes. These controls support Australian ACMA requirements for numbering and calling-party identification, while limiting the impact of a malicious redirect.
Practical checks for Australian deployments
Test one call end to end rather than merely confirming that the forwarding number is saved. Exercise a busy fixed line and PBX extension, verify the second IAM route and charging record, and check the ACM, ANM, REL and RLC sequence in carrier logs. Calls to Australian mobile, 1300 and 1800 destinations can follow different screening and routing paths.
Signalling transport affects resilience. A comparison of TDM and IP backhaul explains why IP/MPLS route diversity, retransmission behaviour and separation of voice media need validation before CFB goes live. Test failover during setup and release so a failed primary path cannot strand the original leg.
The dependable configuration diverts only a confirmed busy call, preserves the +61 destination format, respects Australian numbering and calling-ID obligations, and releases both legs cleanly. That end-to-end discipline gives the caller a transparent service while keeping signalling and carrier charging under control.
