Understanding SS7 Timers in the ISUP Call Release Phase (T1–T20)
In an ISDN User Part (ISUP) call, timers control how long a signalling point waits for the next expected message. They prevent a circuit from remaining reserved indefinitely when a remote exchange, trunk, or signalling link stops responding.
The release phase is especially important. A call may end normally after an answer, fail during setup, or be cleared because of congestion, routing problems, handset behaviour, or an interworking fault. Messages such as REL, RLC, RSC, IAM, ACM, CPG and ANM are interpreted alongside timer expiry.
T1–T20 should not be treated as a single universal stopwatch list. Their meanings and default durations can differ between ITU-T recommendations, national variants, vendor platforms and carrier profiles. Australian networks can also contain legacy TDM exchanges, SS7 gateways and SIP interworking equipment in the same call path.
For engineers working in Sydney, Melbourne, Brisbane or regional networks, the practical task is to identify the timer that expired, the message that should have arrived, and the circuit state at that moment. This is more useful than changing a timeout simply because a call trace looks slow.
How timers fit the ISUP state machine
An outgoing call commonly begins with an Initial Address Message (IAM). The originating exchange then expects progress, such as an Address Complete Message (ACM), Call Progress (CPG) or Answer Message (ANM). If the destination answers, the call enters the speech or bearer phase; if it fails, a Release (REL) message carries a cause value.
During clearing, the exchange that receives REL should release the bearer and return a Release Complete (RLC). The RLC confirms that the remote side has cleared its part of the circuit. If it does not arrive, a release supervision timer protects the local switch from waiting forever.
Timer expiry normally triggers a defined action: sending another release-related message, marking the circuit unavailable, returning a cause such as temporary failure, or raising an alarm. It is therefore a protocol recovery mechanism, not merely a performance setting.
Reading T1 through T20
The early timers in the T1–T20 range generally supervise call establishment and progress. Depending on the implementation, they cover waiting for ACM, ANM or CPG, controlling post-progress waiting, and handling an IAM that receives no valid response. A slow mobile termination, an inter-exchange route problem or an absent backward message can produce similar symptoms.
Later timers are more closely associated with clearing, reset and circuit supervision. Some monitor the response to REL or RSC, while others apply to unusual sequences, continuity checking, blocking, or exchanges between interconnected signalling points. The exact allocation must be checked against the switch vendor’s ISUP timer table.
A useful diagnostic reference is the backward call setup analysis, particularly when ACM, CPG and a cause indication appear in an unexpected order. The trace should show message direction, circuit identification code, point codes, timestamps and the timer state at each transition.
Why release faults are easy to misread
A missing RLC does not always mean that the far-end switch is broken. The REL may have been discarded by an SS7 link, translated incorrectly at a SIGTRAN gateway, or sent on a different circuit identity. MTP3 congestion, link changeover and an overloaded signalling transfer point can delay or lose a message before ISUP processes it.
Cause values add important context. Cause 16 indicates normal call clearing, while causes such as 17, 18, 19, 34 and 41 point towards user busy, no response, no answer, no circuit available or temporary failure scenarios. A timer expiry combined with cause 16 suggests a different investigation from an expiry following cause 41.
Australian operators often carry calls across mixed fixed, mobile and VoIP domains. A call from a Perth business line to a Melbourne mobile may pass through TDM-to-SIP interworking, while emergency and wholesale traffic can follow separate routing policies. National numbering, lawful-interception obligations and carrier-grade monitoring also mean that a local timer adjustment can have operational and compliance consequences.
Setting values without hiding the fault
Timer values should be aligned with the carrier’s signalling profile and the expected network diameter. Increasing a release timer may reduce false alarms on a long or congested route, but it can leave circuits tied up for longer. Reducing it can clear resources quickly while causing valid delayed responses to be classified as failures.
Engineers should compare timer expiry with round-trip signalling delay, link utilisation, retransmission records and switch CPU load. Packet captures from SIGTRAN, MTP3 link statistics and ISUP call-detail records help distinguish a genuine remote delay from a local queueing problem.
The Australian market also includes regional links where latency and backhaul behaviour differ from metropolitan services. A timer profile suitable for a dense Sydney exchange may be inappropriate for a remote Queensland route. Changes should be tested during a controlled maintenance window and reviewed against abandoned calls, blocked circuits and repeat release attempts.
A practical reference for T1–T20
The following summary describes common operational groupings rather than prescribing one vendor’s values. Q.764, national options, carrier documentation and platform configuration remain authoritative for the precise function and duration.
| Timer range | Typical supervision area | Evidence to check | Common response |
|---|---|---|---|
| T1–T4 | Initial backward response, ACM, CPG or ANM progression | IAM, ACM, CPG and ANM timestamps | Clear the call or report setup failure |
| T5–T8 | Answer supervision, delayed progress and early clearing | Call state and cause value | Release the circuit or continue state handling |
| T9–T12 | REL, RLC, RSC and circuit recovery | REL/RLC sequence and CIC status | Mark the circuit released, blocked or suspect |
| T13–T16 | Reset, blocking and abnormal message sequences | RSC, BLO, BLA, UBL and UBA traces | Reset or isolate the affected circuit |
| T17–T20 | Extended supervision and network-specific procedures | Vendor logs, continuity results and link alarms | Apply profile-specific recovery and alarm actions |
A disciplined investigation starts with the first missing expected message, not the final alarm. Record the circuit identification code, point codes, signalling link, cause indicators and elapsed time, then compare the event with the configured T1–T20 profile. The practical takeaway is simple: use timer expiry to locate the broken state transition, verify it against the Australian carrier profile, and change values only after the signalling path has been proven healthy.
