From TDM to SIGTRAN in Legacy Signalling Networks
Telecommunications networks built around time-division multiplexing (TDM) still carry important voice and signalling traffic. ISDN, SS7 links and traditional PSTN exchanges may be reliable, yet their specialist hardware, leased transport and limited flexibility make long-term support increasingly difficult.
The move to SIGTRAN replaces dedicated TDM signalling paths with SS7 protocols transported across IP. M2PA, M2UA, SUA and SCTP allow operators to preserve signalling functions while gaining the routing, resilience and operational visibility associated with packet networks. The transition is an engineering programme rather than a simple equipment swap.
For engineers building their knowledge base, SS7 Training provides useful reference material on ISUP, MTP3, call routing, timers and signalling architecture. These fundamentals remain relevant when a legacy exchange connects to an IP-based signalling transfer point.
Australian operators also need to account for long distances, regional service areas and a concentrated market dominated by large carriers such as Telstra and Optus. A design that works between Sydney and Melbourne may need different resilience assumptions for remote communities, mining regions or routes affected by floods and bushfires.
Why legacy signalling estates persist
TDM networks survive because they are predictable. E1 and STM-1 circuits provide fixed bandwidth, established timing and familiar fault-management procedures. Many switches, media gateways and service platforms still depend on ISUP messages for call setup, release, charging information and number portability procedures.
The commercial difficulty is that spare parts, vendor expertise and circuit provisioning become less available over time. An organisation may also be paying for private transport links whose capacity is reserved even when traffic is low. Migration therefore needs to preserve proven call behaviour while reducing dependence on specialised transmission infrastructure.
| Architecture | Main transport | Strengths | Migration concerns |
|---|---|---|---|
| Traditional SS7 over TDM | E1, T1 or higher-rate circuits | Predictable timing and mature operations | Fixed capacity and ageing equipment |
| SIGTRAN overlay | SCTP over IP with adaptation layers | Reuses SS7 logic and adds path diversity | Requires IP engineering and security controls |
| Native IP voice core | SIP and IP service platforms | Flexible scaling and software integration | Greater redesign and interworking complexity |
Build a complete dependency map
Begin with an inventory of signalling points, point codes, linksets, link utilisation, routes and connected applications. Include service control points, number portability databases, prepaid platforms, lawful-interception interfaces, voicemail systems and international gateways. A diagram that shows only exchanges and signalling transfer points will miss many operational dependencies.
Trace the message flows for ordinary calls, mobile handover, emergency services, short message delivery and failed-call scenarios. In Australia, emergency call handling and service continuity across wide geographic areas deserve explicit testing. Document which systems expect MTP2, MTP3, SCCP, TCAP or ISUP behaviour, and record every timer and retry assumption.
Capacity planning should cover busy-hour attempts, bursts, link occupancy and signalling message size. The link dimensioning guide can support an Erlang B-based assessment, although IP transport planning must also consider SCTP associations, packet loss, jitter and failover overhead.
Select a staged migration path
A common approach is an overlay deployment. New SIGTRAN signalling gateways connect to existing TDM signalling points, allowing selected traffic to move while the original links remain available. This provides a controlled rollback path and lets engineers compare message traces across both environments.
Another option is gateway replacement, where a legacy signalling transfer point is retired after its peers and applications have been migrated. This can reduce complexity faster, but it creates a larger change window. A hybrid approach is often suitable: retain TDM for high-risk or remote routes while moving stable national and metropolitan traffic to IP.
Migration waves should be based on service boundaries rather than equipment age alone. For example, an operator might begin with a low-risk inter-exchange route in Melbourne, then extend the design to Sydney and regional hubs after observing live performance. Each wave should have entry criteria, rollback triggers and a named owner for every dependent platform.
Design resilience into SIGTRAN
SCTP multihoming is central to a robust SIGTRAN deployment. Separate network paths, interfaces, routers and power domains reduce the chance that one local failure interrupts signalling. The IP underlay should use diverse fibre routes where practical, particularly across large metropolitan aggregation points and long-haul links to regional sites.
Keep signalling traffic separate from general enterprise data. Dedicated VRFs, carefully controlled routing, traffic classification and deterministic firewall policies help prevent congestion or accidental exposure. Latency targets should be defined for each signalling relationship, with alarms for packet loss, association changes, retransmissions and abnormal failover frequency.
Testing must include link failure, routing convergence, gateway restart, loss of a signalling processor and partial site isolation. Verify that calls are released correctly when transactions cannot complete; silent message loss can create stuck resources and misleading capacity reports. Capture baseline measurements before migration so that the IP design is compared with actual service behaviour rather than assumptions.
Migration controls that reduce risk
Operational discipline matters as much as protocol knowledge. Maintain a version-controlled signalling matrix, approved change scripts and a current contact list for carriers, vendors and internal service owners. Schedule high-impact work outside Australian peak calling periods and avoid major cutovers during known disaster-response or public-event windows.
Useful controls include:
- Run parallel TDM and SIGTRAN paths until service and fault metrics meet agreed thresholds.
- Validate point codes, routing keys, global titles and translation tables before traffic is enabled.
- Monitor call completion, release causes, SCCP failures, SCTP associations and retransmission rates.
- Test security controls against scanning, spoofed signalling endpoints and unauthorised route changes.
- Keep a rehearsed rollback procedure that can be completed within the permitted maintenance window.
Training should cover packet capture, MTP3 route management, SCTP diagnostics and application-level ISUP analysis. Teams accustomed to physical link alarms may otherwise miss an IP-layer fault that presents as a signalling congestion problem.
Turn the strategy into a controlled cutover
A successful programme defines the target state before selecting products. Decide which SS7 functions will remain, which will be virtualised, where signalling gateways will sit, and how the network will interwork with SIP, mobile cores and external carriers. This prevents a temporary overlay from becoming an undocumented permanent architecture.
Governance should include security review, operational acceptance, vendor interoperability testing and a post-cutover service review. Retiring TDM circuits only after several stable billing and fault-reporting cycles reduces the risk of removing a hidden dependency. Record lessons from each migration wave and apply them to the next route or service group.
The practical next step is to create a current-state signalling inventory containing every linkset, point code, route, application dependency and busy-hour measurement, then use it to select one low-risk pilot link for a documented SIGTRAN overlay.
