Real-time VoIP fraud detection
Catch fraud calls before they connect.
VoIP Radar evaluates each call the instant it reaches your network. Spoofed caller ID, robocall bursts, IRSF, Wangiri and toll-fraud attempts are scored and stopped in-line — so your subscribers, margins and reputation never feel them.
Detection coverage
What VoIP Radar picks up
Each call is checked against signaling, numbering, reputation and behavioral signals at once. These are the threats carriers ask us to stop most often.
Spoofed caller ID
Attestation results, numbering validity and per-customer number inventories expose calls presenting numbers the caller has no right to use.
Robocall campaigns
Calls-per-second spikes, short-duration clusters and low answer rates reveal auto-dialer traffic within seconds of it starting.
Suspicious routing
Traffic that loops, laundered through unknown upstreams or arriving from high-risk regions is risk-scored at ingress.
Bad reputation
Shared threat intelligence, complaint signals and historical outcomes feed a continuously updated number and source reputation.
IRSF & Wangiri
Calls toward premium-rate ranges and one-ring fan-out patterns are flagged before revenue is lost or subscribers are baited.
Toll fraud & PBX abuse
Hijacked extensions, registration brute force and unusual international dial patterns are shut down at the edge.
Under the hood
How in-line blocking works
When a SIP INVITE arrives, VoIP Radar reads the elements that matter — calling and called numbers, Identity header and attestation, the customer and trunk it came in on, routing history, timing and destination — and enriches them with that customer's historical baseline and network-wide reputation data.
A machine-learning risk model turns those signals into a single score in milliseconds. Your policy decides what happens next: allow, allow and tag, rate-limit, or reject with the correct SIP response. High-risk events raise an alert to your NOC with the evidence attached.
Every decision is written to a tamper-evident audit log alongside the reasons behind it, so you can answer a traceback request, a customer dispute or a regulator with facts rather than assumptions.
Policy actions you control
Tune thresholds per customer, trunk or product line — strict on new accounts, lighter on long-standing enterprise traffic.
| Risk level | Typical signals | Default action |
|---|---|---|
| Low | Known customer, verified numbers, normal cadence | Allow and sign with the correct attestation |
| Elevated | New destination mix, rising CPS, unverified BYOC numbers | Allow, tag, alert and downgrade attestation |
| High | Invalid or DNO calling number, IRSF range, wangiri fan-out | Reject with SIP 603 / 608 and alert NOC |
| Critical | Registration brute force, sustained fraud from one account | Block source, suspend account, open incident |
Fraud detection FAQ
Does VoIP Radar add latency to calls?
Scoring runs in milliseconds and is designed to sit in the signaling path without a noticeable effect on post-dial delay. Legitimate calls continue as normal.
Do we need to replace our switch or SBC?
No. VoIP Radar works alongside the softswitch and SBC you already run — commonly as a SIP redirect/route server or via API from your routing logic.
How are false positives handled?
You set the thresholds and actions. Elevated-risk calls can be tagged or rate-limited instead of blocked, and every decision carries a reason code so your team can review and whitelist quickly.
Can it help with traceback requests?
Yes. Because every call decision is logged with the customer, numbers, attestation and score, you can locate a traced call and respond with evidence well inside the 24-hour window.
Get in touch
Find out where your network stands
Tell us about your switch and traffic. A VoIP Radar engineer will map your STIR/SHAKEN, RMD and fraud exposure and send a written assessment — free, within one business day.
- STIR/SHAKEN & attestation review
- RMD certification gap check
- Fraud & robocall exposure snapshot
- Deployment plan for your SBC / softswitch