load sharing with point code pairs and SLS masking techniques
In Signaling System No. 7 networks, the Signal Transfer Point handles millions of messages per second across interconnected carriers. When a pair of STPs is deployed in parallel, point code pairs combined with Signalling Link Selection masking become the primary mechanism for distributing load evenly between them. Understanding how SLS bits are extracted and applied helps engineers design redundant signalling paths that survive link failures without congesting a single node.
Australian operators such as Telstra and Optus run large mated-pair STP configurations across capital cities like Sydney, Melbourne, and Brisbane, where regulatory requirements from the ACMA push for geographical redundancy. Load sharing matters especially in regional areas such as Perth and Adelaide, where long-haul signalling links must be balanced carefully to prevent one side of the mated pair from becoming a hot spot during peak traffic.
The technique of SLS masking uses a controlled subset of the SLS bits, derived from the MTP3 routing label, to decide which physical link within a linkset, or which STP within a mated pair, carries a particular message. Designers trade off granularity against administrative overhead when choosing how many bits to mask, and the rest of this article examines those choices in practice.
How point code pairs work in STP load sharing
A point code pair consists of two destination point codes assigned to physically distinct STPs that share the same logical role within the network. When a signalling point chooses a route, the STP selection logic looks at a small portion of the SLS field to decide between the two members of the pair, producing balanced traffic distribution as long as calling patterns generate uniform SLS values.
Carrier networks in Sydney and Melbourne typically run multiple point code pairs to isolate different traffic classes, such as ISUP voice traffic separated from SCCP-based intelligent network queries. Granularity depends on how many SLS bits are used to index the available links and how the mask is configured on each STP.
For engineers familiar with intelligent network subsystems, SCCP subsystem numbers explained interact with the same SLS bits that drive load sharing, which is why masking decisions affect more than voice call routing.
SLS field structure and bit selection
The SLS field is part of the MTP3 routing label and, depending on the variant, occupies between four and eight bits of the message. The lower bits are usually the most significant for load sharing because they change most frequently between consecutive messages on the same call, spreading traffic evenly across available links.
Bit selection in a typical mated-pair STP configuration uses either the two least significant bits for four-way sharing or the three least significant bits for eight-way sharing. The choice influences how evenly the load is distributed when upstream traffic does not exhibit uniform randomness, which is common when certain destinations dominate the signalling mix.
Australian operators with large interconnect volumes at exchanges in Brisbane and Perth sometimes observe that certain SLS values are over-represented because of how their vendors encode the Calling Party Number fields. Adjusting the mask to use bits that vary more reliably in those contexts often produces smoother distribution than the default choice.
Comparing mask lengths and traffic distribution
| Mask length | Bits used | Logical links supported | Typical use case | Distribution quality |
|---|---|---|---|---|
| 1 bit | SLS[0] | 2 | Small mated pairs | Adequate for low traffic |
| 2 bits | SLS[0–1] | 4 | Standard voice mated pair | Good for most networks |
| 3 bits | SLS[0–2] | 8 | Mixed ISUP and SCCP | High for diverse traffic |
| 4 bits | SLS[0–3] | 16 | Dense carrier cores | Excellent when uniform |
| Full SLS | All bits | 16–256 | Special routing cases | Best when available |
The table summarises the most common masking configurations. Short masks are easier to administer but offer less granular distribution, while longer masks require more linksets and administrative tracking to realise the potential gain.
Implementation considerations for Australian networks
Regulatory and operational realities shape how SLS masking is rolled out across Australian networks. The NBN's multi-technology mix and the growing volume of mobile-originated signalling from carriers in Adelaide and regional Queensland both create patterns that an unconfigured mask can struggle to balance. Engineers typically perform trace captures on production STPs during busy hour to confirm that masking decisions produce the expected distribution.
Hardware platform limits also matter. Older STP equipment in some regional exchanges supports only certain mask widths, and upgrading requires coordination between vendor support contracts and ACMA reporting for any service-affecting change. Planning the migration in advance avoids the situation where one half of a mated pair absorbs all traffic during a cutover.
Network teams sometimes coordinate masking choices with unrelated infrastructure projects to reduce overhead during maintenance windows. While planning downtime, engineers often consult vendor bulletins and even resources that discuss broader topics like reliable online casino guides during downtime research, which highlights how diverse the maintenance reading list can become.
Practical recommendations for STP masking deployment
- Begin with a 2-bit mask for new mated pairs and only increase to 3 or 4 bits when trace data shows uneven distribution.
- Verify SLS bit randomness by sampling live traffic at peak hour rather than relying on vendor defaults.
- Document the mask configuration alongside the point code pair table to simplify audits during ACMA reviews.
- Test failover behaviour by failing individual links and confirming that remaining linksets absorb the displaced load without exceeding 70 percent utilisation.
- Review masking choices whenever a new service such as VoLTE or 5G non-standalone signalling is introduced.
What stays with the reader after working through this material is the link between mask length, SLS bit variability, and real-world traffic patterns. Australian networks carry enough concentrated signalling around the eastern seaboard and the long routes to Perth that masking decisions made during design directly shape how a mated pair behaves on the busiest day of the year. Choosing the mask deliberately, validating it with captured traces, and revisiting it as new services appear keeps load sharing predictable, auditable, and resilient when a single link or STP fails.
