SS7 signaling inspection
One firewall on the signaling path. MAP, CAP and SCCP are decoded, normalized and scored against a single policy engine before any message reaches your SMSC, HLR or MSC.
BroadNet's SS7 firewall stops signaling fraud and grey-route A2P traffic for mobile operators by inspecting SS7 signaling in-line, before it reaches the core network. An SS7 firewall with in-line MAP/SCCP filtering between the international STP and your core network. Sender ID control, route validation, SRI-SM cache filtering, content filtering and A2P monetization — all from a single signaling-layer firewall. Deployed at Tier-1 operators across the MENA region.
One firewall on the international signaling path. Every MAP, CAP and SCCP unit is decoded per-IE, categorized against GSMA FS.11, and a verdict is returned in under five milliseconds. No re-route, no broken roaming.
International SS7 (SIGTRAN, M3UA) links inbound from roaming partners and grey routes.
MAP, CAP and SCCP per-IE decode. GT, AC, CdPA/CgPA, IMSI, MSISDN normalized for policy lookup.
Category 1/2/3 classification, SRI-SM cache validation, sender-ID and route checks, velocity buckets.
Allow · modify · quarantine · drop · monetize. Per-route, per-partner, per-subscriber. Maker-checker promoted rules.
Only validated signaling reaches the home network. Quarantined messages mirrored for fraud team review.
International SS7 (SIGTRAN, M3UA) links inbound from roaming partners and grey routes.
MAP, CAP and SCCP per-IE decode. GT, AC, CdPA/CgPA, IMSI, MSISDN normalized for policy lookup.
Category 1/2/3 classification, SRI-SM cache validation, sender-ID and route checks, velocity buckets.
Allow · modify · quarantine · drop · monetize. Per-route, per-partner, per-subscriber. Maker-checker promoted rules.
Only validated signaling reaches the home network. Quarantined messages mirrored for fraud team review.
Decoded at the STP, scored in real time, and acted on in under five milliseconds. One signaling-layer engine handles MAP, CAP and SCCP; one policy DSL writes the rules; one audit store keeps the evidence.
One firewall on the signaling path. MAP, CAP and SCCP are decoded, normalized and scored against a single policy engine before any message reaches your SMSC, HLR or MSC.
Out-of-the-box rule packs aligned with GSMA FS.11 for SS7 interconnect. Category 1 never crosses the network boundary; Category 2 is validated against home-network origin; Category 3 is velocity-scored per subscriber.
Every message is fingerprinted across origin GT, SCCP CdPA/CgPA, application context and historical pattern. Anomaly scores feed a policy DSL that decides allow, modify, quarantine or drop in a single hop.
AnyTimeInterrogation, ProvideSubscriberInfo, ProvideSubscriberLocation and SendRoutingInfo abuse are caught at first contact. Foreign GTs probing for IMSI–MSC–VLR tuples are blocked before a single subscriber is geolocated.
MT-ForwardSM is validated against a live SRI-SM cache by IMSI / MSC / SMSC tuple. Spoofed sender IDs, A2P leakage onto grey routes and SIM-swap reconnaissance are stopped before the SMSC sees them, not after.
Every signaling decision is committed to a tamper-evident audit store with HMAC-signed event chains. Regulator-ready exports, role-based dashboards and incident replays — your NOC sees what the firewall saw, frame-for-frame.
The five signaling attack classes that account for most subscriber harm and most operator revenue loss — and the layer at which the firewall catches each one.
Originator addresses fabricated to impersonate banks, brands and government services. Caught by SRI-SM cache mismatch and sender-ID registry cross-check.
Foreign GTs query the HLR to triangulate a subscriber's serving cell. Blocked at the network boundary as GSMA FS.11 Category 1 messages.
Attackers probe for IMSI–MSC bindings ahead of social-engineered swaps. Pattern-scored per origin, then quarantined for fraud-team review.
International revenue-share fraud, Wangiri loops and spoofed CLIs intercepted at the signaling layer before they ring the subscriber.
Unmonetized enterprise traffic injected via roaming partners. Forced onto the agreed commercial route — or blocked, with the revenue recovered.
Two MAP operations explain a disproportionate share of the subscriber-privacy and OTP-interception incidents reaching regulators today. The firewall handles both at the international edge — categorised, scored and signed — before they reach a home HLR or SMSC.
Foreign HLR probes that resolve to a serving cell ID, used by surveillance operators and abusers to triangulate a subscriber's real-world position. The firewall classifies these as GSMA FS.11 Category 1 and drops them at the network boundary, with a signed audit entry for the regulator file.
SRI-SM queries that map a subscriber to their serving MSC/SMSC, the reconnaissance step before MT-ForwardSM hijack of OTP and 2FA traffic. The firewall validates each query against a 24-hour SRI-SM cache; cache-miss queries from foreign GTs quarantine for fraud-team review.
The signalling layer needs protocol enforcement, commercial policy, and behaviour-aware adaptation in the same engine. The three pillars run as one decision path — sub-5 ms inspection latency per signalling unit — so the right verdict reaches the right network element on the first hop.
Protocol-level enforcement at the signaling layer. MAP operation allow/deny, Global Title filtering, OPC/DPC point-code controls, MCC/MNC conditions per partner. Every malformed or out-of-context signaling unit drops at decode.
Commercial and operational policy expressed as data. Route selection per partner agreement, product-class policies (OTP, transactional, promotional), monetisation of bypass traffic toward agreed A2P channels. The firewall enforces the contract, not just the spec.
Behaviour-aware enforcement that adapts to live traffic. TPS thresholds, burst detection, time-of-day windows, velocity per origin GT. The dynamic layer catches what static rules cannot — emerging campaigns, new SIM-box farms, sudden roaming anomalies.
FS.11 categorisation expressed as readable YAML. MAP and CAP operations named, not numbered. Sender-ID, route and A2P rules promoted through maker-checker — versioned, signed, auditable.
# /etc/broadnet/ss7fw/rules.d/fs11-category1.yaml# GSMA FS.11 Category 1 — never crosses the network boundaryrule: cat1_ati_block match: application_context: anyTimeInterrogation origin: !is_home_network action: drop audit: regulator.signal_event fs_tag: FS.11-CAT1 rule: cat1_psi_block match: application_context: provideSubscriberInfo origin: !is_home_network action: drop fs_tag: FS.11-CAT1 # Category 2 — origin must be validated against home-network MAPrule: cat2_sri_sm_validate match: application_context: sendRoutingInfoForSM validate: msc_id: from sri_sm_cache(ttl=24h) on_mismatch: action: quarantine notify: fraud-ops@operator.tldInline between the international STP and your core. Every message passes the firewall before it reaches an HLR, SMSC or MSC — and every decision is streamed to your NOC in real time.
Active-active pair sits on the international signaling path. Decisions are enforced in real time — block, allow, modify, monetize.
Mirror feed from the STP for baseline establishment and analytics. Move to in-line enforcement when policy and false-positive rates stabilize.
International edge runs in-line, domestic core runs in detect-and-alert. Common policy plane, common audit store, one operator dashboard.
GSMA FS.11 categorization for SS7 interconnect is enforced by default and maintained by the BroadNet signaling security team, with regulator-ready exports on demand.
Yes. The firewall presents as an SCCP/SIGTRAN signaling peer and integrates with the common STP platforms operators run today, across mixed vendor stacks. No STP routing-table surgery is required beyond redirecting the international point codes through the firewall pair.
We almost always start with an out-of-band mirror from the STP for 14–30 days. That gives the security and RA teams a clean baseline of the international traffic and lets us tune the FS.11 rule pack against the operator's actual roaming-partner mix. Once the false-positive envelope is agreed in writing, the firewall flips to inline active-active. Hybrid (inline at the international edge, mirror at the domestic core) is the common steady state.
BroadNet's signaling security team publishes a versioned rule pack every quarter, with hot-fix patches for new threat intelligence pushed in days. Each rule pack ships with the IR.71, IR.81, and IR.82 conformance test packs so your team can run regression before promoting it. The operator pins or vetoes any version from the dashboard — nothing is auto-promoted to production.
Per inspected signaling unit (MAP/CAP), banded by monthly volume. No per-link, per-point-code or per-Erlang fee. There is a one-time NRE for the STP integration and a small annual maintenance fee covering rule-pack updates, the threat-intel feed, and quarterly architecture reviews. Operators typically see payback from recovered A2P revenue and avoided fraud loss.
Yes. The policy DSL is the same one our security team uses internally. Operators routinely write rules to enforce their bilateral roaming agreements, to whitelist authorized law-enforcement location queries, to monetise specific bypass flows rather than drop them, and to feed verdicts to in-house SIEM and fraud platforms. Custom rules go through the same maker-checker promotion as ours, with full audit trail.
TTL is 24h by default and tunable per operator. The cache is populated from observed legitimate SRI-SM and refreshed on every successful MT-ForwardSM. We expose hit rate, miss rate, and quarantine rate per partner GT so RA teams can spot a sudden cache miss spike (usually the first sign of a spoofing campaign or a partner re-keying their SMSC GT).
Yes. Every decision is HMAC-signed, written to a tamper-evident store, and retained for the operator's regulatory window. We ship pre-built export templates for TRC (Jordan), CST (KSA), NCC (Nigeria), TDRA (UAE), CRA (Qatar), TRA (Oman), TRA (Bahrain), ARPT (Algeria), ANRT (Morocco), and INT-T (Tunisia). The export module also produces the FS.11 conformance manifest that several MENA regulators now ask for at licence renewal.
In steady state, false positives against bona-fide roaming partners are very low. We measure the rate weekly and expose the per-partner number in the dashboard, so RA teams always see it. The mirror-mode bake-in phase exists exactly so that this envelope is agreed and signed off before enforcement goes live.
Need the FS.11 conformance manifest or a working session with our signaling engineers? Request signaling security review
Application-layer SMS firewall sitting on the SMPP interface to your SMSC.
Vendor-agnostic capture and correlation for SS7, SIGTRAN and SIP — the companion intelligence layer.
Hosted SMSC with multi-routing, multi-language and unified billing.
Real-time observability across firewall, SMSC, signaling and routes.
Schedule an operator briefing. We'll walk the FS.11 categorization, the SRI-SM cache, the A2P monetization model, and the deployment plan for your network — backed by 24/7 BroadNet NOC support.
Resources
SS7 Firewall datasheet
Go deeper on the concepts behind this product.