SS7 firewall rule examples for blocking unauthorised MAP operations

Unauthorised MAP activity is most effectively addressed through peer identity, operation context and transaction state. Mobile core networks use MAP for subscriber location, routing and call-control procedures, so a blanket block can create signalling failures beyond the suspicious activity it was meant to stop.

The examples here describe defensive screening within a carrier-controlled environment backed by laboratory instrumentation and change controls. Rules should be expressed in vendor-neutral terms before implementation in an SS7 firewall, SIGTRAN gateway or specialised signalling monitor, because deployment capabilities differ significantly.

Where application screening belongs

The screening decision needs to sit above basic message transport. MTP3 and SCCP checks can constrain routing and endpoint access, while operation-level controls inspect the MAP service carried inside TCAP. Treating every unknown operation as harmful removes useful diagnostic evidence and may interfere with roaming.

Control layer Example rule Primary benefit Main caution
Peer and route admission Accept links only from approved STP neighbours Blocks traffic from unapproved peers Fail-over routes need explicit approval
MAP operation screening Permit MAP_UPDATE_LOCATION only between approved VLR and HLR roles Limits false location registrations Roaming behaviour varies by vendor
Destination and parameter checks Reject unexpected SPC or GT values and malformed lengths Rejects malformed or misdirected transactions Routing data must remain synchronised
Request-rate controls Throttle repeated MAP_CANCEL and routing requests Limits flooding and retry storms Thresholds require per-peer tuning

Before a rule is enabled, identify whether the traffic belongs to a domestic HLR–VLR relationship, a partner interconnect or an internal service. The same operation can be legitimate from one peer and anomalous from another. The examples use symbolic operation names and service roles rather than device-specific configuration syntax.

Representative rules for a controlled lab

Location-update control is a useful starting point. A rule that permits MAP_UPDATE_LOCATION only between approved VLR and HLR endpoints rejects that operation from an unrelated peer without banning it across the network. This preserves expected registration behaviour while exposing attempts to insert false subscriber state.

A second pattern permits MAP_UPDATE_GMSC only when the source is an approved serving or gateway MSC and the destination is an assigned HLR. A third allows MAP_SEND_ROUTING_INFO from approved gateways, while rejecting unexpected source identities and destination scopes. These are role-based allow rules with explicit deny behaviour rather than broad blocks on operation names.

Volumetric controls provide another layer. Repeated MAP_CANCEL or routing requests can consume transaction resources even when their format is valid. Apply per-peer rate thresholds, correlate bursts with transaction state and define whether excess traffic is delayed, challenged or dropped.

Limits of operation-based filtering

SS7 is a protocol family, not a single network port. An IP firewall that only filters IP addresses cannot identify a MAP operation inside MTP3, SCCP and TCAP. On SIGTRAN deployments, screening may need to decode M3UA and M2PA traffic, so transport support and inspection depth must be confirmed before a policy is written.

Rejecting an operation is also weaker than authenticating its origin. Controls should combine peer allow-lists, routing consistency, transaction correlation and monitoring for attempts that misuse permitted operations. Where deployment supports cryptographic protection of the transport, configure it as part of the trust model rather than assuming a valid source address proves legitimacy.

Timeout behaviour determines the customer impact of a bad rule. Silent drops can leave callers waiting, while an immediate rejection can reveal the screening decision to a peer. The reference on network timeout effects explains how timeout settings affect SS7 call setup and post-dial delay.

Australian network and legal realities

Australia’s carrier market is concentrated, with Telstra, Optus and TPG Telecom operating major mobile networks and several brands sharing underlying infrastructure. A blunt policy at an interconnect can therefore affect multiple retail and enterprise services. Each rule should be mapped to the actual carrier relationship and network role before it is enabled.

The +61 country code and Australian mobile ranges beginning with 04, including 0400, provide useful normalisation checks. They do not authenticate a sender or establish a legitimate HLR. Emergency routing for 000 also needs a dedicated regression test, because a rule blocking an unfamiliar but valid path can have serious consequences.

Subscriber identifiers and location records can bring the Privacy Act 1988 (Cth) into operational and audit processes. Restrict access to rejected-message payloads and document the purpose of retained data. Because the ACMA regulates the Australian telecommunications sector, changes with service-wide effects also warrant a check against the carrier’s legal and regulatory obligations.

Operational schedules must reflect Australia’s geography. Perth is two hours behind Sydney in standard time, while Melbourne is one hour behind. Rate limits, change windows and partner support coverage therefore need a common time reference rather than an assumed single national working day.

A controlled rollout for Australian operators

Treat every rule as a change to a live dependency. Test it first with a simulator and sanitised transactions, then exercise normal calls, roaming, emergency routing and partner failover. Capture the rejection event and verify that transaction timers release correctly. A policy that works during a quiet test may behave differently during a regional failover or a busy evening in Melbourne.

Roll out new operation rules in alert mode, compare rejections with call-setup success and post-dial delay, and promote only the rule whose expected behaviour has been demonstrated. Keep peer exceptions owned and reviewed, attach expiry dates, and maintain a tested rollback path for every blocking change.

Begin in the laboratory with a single rule allowing MAP_UPDATE_LOCATION only between approved VLR and HLR endpoints, capture one rejected test transaction from an unapproved peer, and enable it in production only after that check and the 000 emergency-routing regression test pass.