SIP-I vs SIP-T: interworking ISUP with next-generation networks
Australia's telecommunications landscape is shifting fast. Telstra has been progressively switching off its legacy PSTN and ISDN services, with copper retirement accelerating alongside the National Broadband Network rollout from suburbs like Parramatta to regional centres such as Toowoomba. As circuit-switched voice gives way to all-IP infrastructures, the signalling systems that carried call setup, teardown, and supplementary services must also evolve. The relationship between ISUP and modern SIP-based architectures becomes critical for engineers maintaining inter-carrier connectivity and operators planning migration paths.
Two protocols dominate this transitional layer. SIP-I and SIP-T both enable ISUP signalling to traverse IP networks, but they approach the task with different philosophies, payload formats, and intended deployments. Understanding where each shines helps Australian carriers, integrators, and regulators like the ACCC ensure voice interoperability continues smoothly during one of the country's largest infrastructure transitions.
Origins of ISUP and the push toward IP
Signalling System No. 7 has governed trunk signalling between exchanges for decades, with ISUP handling the call-control messages that route bearer channels across the PSTN. In Australia, ISUP variants such as ISUP-95 have long been the connective tissue between Telstra's exchanges and competing networks. When voice moved to IP, the industry needed ways to carry these ISUP messages across packet-switched domains without losing fidelity or breaking inter-carrier agreements.
SIP was designed for session establishment in IP-only environments, using headers and bodies to negotiate media sessions. SIP-I and SIP-T emerged as the two main answers to how ISUP semantics could be preserved while leveraging SIP's flexibility. Their differences reflect competing priorities: strict 3GPP-aligned IMS deployments versus pragmatic interworking for legacy trunks.
How SIP-I and SIP-T carry ISUP messages
SIP-T takes a tunneling approach, embedding entire ISUP message bodies inside SIP INVITE and other requests using the MIME multipart/mixed content type. The receiving gateway extracts the ISUP payload and processes it as if it arrived over an SS7 link. This method preserves every ISUP parameter and suits carrier-to-carrier interconnects where both ends want full visibility of the original signalling.
SIP-I, standardised under ITU-T Q.1912.5 and aligned with 3GPP IMS specifications, encodes ISUP messages explicitly using SIP headers and bodies in a structured way. It is the protocol of choice within IMS architectures where strict adherence to 3GPP release specifications matters, and it can also coordinate bearer resources in mixed TDM/IP environments.
Bearer control and media negotiation
A subtle distinction lies in how each protocol interacts with the underlying bearer. SIP-T typically relies on SIP's standard SDP offer/answer mechanism to set up RTP streams, leaving the ISUP content to provide call context but not to manage media directly. This separation suits scenarios where the IP network is purely a transport for what would have been a circuit-switched call.
SIP-I can operate in a more integrated fashion, particularly in environments where circuits and IP paths coexist. Profiles defined for IMS allow SIP-I messages to coordinate bearer resources across heterogeneous network segments, useful during phased migrations where some traffic still flows through TDM switches. Engineers planning hybrid deployments in Brisbane or Adelaide often weigh this capability heavily.
Carrier adoption across Australia
The three largest Australian operators have approached the SIP-I/SIP-T question differently. Telstra, as the dominant wholesaler, has historically favoured SIP-T for intercarrier voice, leveraging its well-established profile to maintain compatibility with international partners. Optus and TPG Telecom have explored SIP-I within their IMS cores, particularly for mobile and enterprise services that benefit from tighter 3GPP alignment.
Before committing to either path, operators typically weigh a handful of practical contrasts. The table below summarises the key differences that shape Australian carrier decisions.
| Aspect | SIP-T | SIP-I |
|---|---|---|
| Standardisation | IETF (RFC 3372 and related) | ITU-T Q.1912.5, 3GPP IMS |
| ISUP representation | Tunneled as MIME body | Encoded in SIP headers and bodies |
| Typical deployment | Carrier-to-carrier interconnects | IMS and 3GPP-aligned networks |
| Bearer control | SIP negotiates separate RTP streams | Can coordinate IP and circuit bearers |
| Parameter visibility | Full ISUP payload accessible | Profile-dependent, often summarised |
The Telecommunications Consumer Protections Code and ACCC determinations around wholesale voice services also shape these decisions. As the NBN reaches more premises, SIP-based interconnect becomes the default, and existing TDM interconnects at sites in Sydney and Melbourne are being phased out in line with Telstra's disconnection schedule, which has moved several milestones forward since 2020.
Migration paths for operators
Several considerations guide Australian operators moving traffic from SS7/M3UA-only setups to environments that include SIP interconnects:
- Existing inter-carrier agreements and the ISUP variant in use
- Presence of legacy switches still requiring circuit-side coordination
- Capacity and redundancy requirements for media gateways
- Compliance with ACCC and ACMA technical standards during transition
Operators with established IMS investments often find SIP-I simpler to integrate, while those maintaining pure IP interconnects with international carriers frequently default to SIP-T. The choice hinges less on technical superiority and more on what the surrounding ecosystem expects.
Reliability and security considerations
Both SIP-I and SIP-T inherit the security challenges of their parent protocols. SS7's historical vulnerabilities, including well-documented signalling attacks, prompted Australian carriers to deploy robust firewalls and screening as recommended by the ACMA. SIP requires careful attention to authentication, encryption, and rate-limiting to prevent toll fraud and enumeration attacks that could otherwise expose subscriber details.
For interconnects carrying sensitive traffic such as emergency calls or financial services, redundancy matters as much as protocol choice. Australian operators commonly deploy geo-redundant session border controllers across data centres in multiple capital cities, with failover paths tested against the realities of long-haul latency between Perth and Sydney.
Choosing the right protocol
Selecting between the two approaches comes down to matching the protocol to the deployment context:
- Use SIP-T when interworking with international carriers or non-IMS SIP networks
- Choose SIP-I when integrating with 3GPP IMS cores or mobile voice services
- Adopt SIP-I where hybrid TDM/IP bearers must be coordinated
- Default to SIP-T for straightforward SIP-to-SIP trunking scenarios
Engineers seeking deeper protocol walkthroughs and configuration examples can review the practical material available through hands-on protocol guides. Many practitioners find that direct exposure to both protocols clarifies their respective strengths far faster than reading specifications alone.
What should stay with you after working through this comparison is that neither SIP-I nor SIP-T is universally better; each serves distinct architectural needs in Australia's evolving voice landscape. SIP-T remains the pragmatic choice for pure IP interconnects, particularly for international gateways, while SIP-I offers tighter IMS alignment for mobile and converged operators. As the NBN-driven PSTN retirement continues and ACCC oversight keeps the transition accountable, the underlying ISUP semantics will be preserved through whichever protocol best matches each carrier's existing infrastructure and future direction.
