Architecture

Network-Level DDoS
Mitigation & Monitoring

CoreDetection analyses live network traffic from flow telemetry to detect DDoS attacks in near real time and triggers mitigation actions based on operator-defined rules and workflows. A 3-layer Smart Detection engine scores every candidate event — rhythm, fingerprinting, and attack memory — then applies ASN trust, per-prefix baselines, and consensus weighting before BGP or alerts fire. Engineers retain full control over how, when, and where response is applied.

Detection Flow
Routers
Flow Export
CoreDetection
Smart Detection
BGP / Alerts
Response
End-to-End Flow

From Flow Export to Auto-Response

01

Routers Export Flows

Your border routers send NetFlow, IPFIX, or sFlow directly to CoreDetection. No inline tap, no packet-path changes — out-of-band deployment on commodity hardware.

NetFlow v9IPFIXsFlow
02

Smart Detection Analysis

Thresholds surface candidate events. The 3-layer engine scores each one — rhythm, fingerprinting, attack memory — then applies ASN trust, baselines, and consensus weighting.

3 LayersSmart DetectionBehavioral
03

Dashboard & Reports

Every confirmed event appears on your operations dashboard with rhythm, fingerprint, and memory score breakdown. Attack history is stored and exportable as JSON or CSV reports.

Live DashboardForensics
04

BGP / Alerts Response

Confirmed attacks trigger BGP blackhole, FlowSpec (drop/rate-limit/redirect/mark), or IP blocklist enforcement automatically. Telegram, email, and webhooks fire simultaneously — all logged in your dashboard.

BGPFlowSpecBlocklistAlerts
DDoS Detection

How CoreDetection Detects DDoS Attacks

CoreDetection ingests traffic telemetry from routers, switches, and virtual networks using standard, widely supported methods. Thresholds surface candidate events; Smart Detection then scores each one through three analysis layers with supporting signals for ASN trust, baselines, and pseudo-L7 inference.

Detection thresholds are fully configurable, allowing engineers to tune sensitivity based on network characteristics and historical attack patterns. No packet payload inspection — scalable and vendor-agnostic.

Telemetry Sources

NetFlow v9IPFIXsFlow v5SPAN / port mirroring

Metrics Monitored

  • Bandwidth utilisation
  • Packets per second (PPS)
  • Flow count
  • Smart Detection score (0–100)

Near-Instant Detection

sFlow and SPAN / port mirroring 1–2 seconds
NetFlow v9 and IPFIX 5–30 seconds
Smart Detection scoring <60 seconds

Attack Types Detected

  • UDP, TCP and ICMP flood attacks
  • SYN, SYN-ACK and FIN floods
  • Fragmentation-based IP protocol attacks
  • DNS, NTP, SSDP, SNMP amplification
  • Multi-vector attacks combining multiple techniques
Smart Detection Engine

3 Analysis Layers
+ Supporting Signals

The engine runs three independent layers on every candidate event, then applies consensus weighting, ASN trust, baselines, and pseudo-L7 inference to produce a final score 0–100. Score ≥70 confirms an attack.

01 Layer 1

Rhythm Analysis

Does this traffic move like an attack?

Analyses 61-second traffic samples for variance, smoothness, and burst patterns. Erratic spikes score high; gradual legitimate ramps score low.

02 Layer 2

Traffic Fingerprinting

Do these sources look like a botnet?

Maps source IP clustering, ASN diversity, and geographic dispersion. Includes trusted-ASN score adjustment to reduce false positives on CDN and peer traffic.

03 Layer 3

Attack Memory

Have we seen this attack before?

Stores signatures of confirmed attacks (up to 1,000). New events matched at 85%+ similarity get instant high-confidence recognition.

Supporting Signals

Applied after the three core layers — these refine the final score but are not separate analysis layers.

Weighted Consensus

Weighted scoring across rhythm, fingerprint, and memory — strong signals (≥70) override weak ones instead of a flat average.

Trusted ASN Override

Operator-configured trusted ASNs can veto false positives when traffic is predominantly from verified legitimate sources.

Behavioral Baselines

Per-prefix exponential moving average learns normal Gbps, PPS, and source counts — reduces score for in-range traffic, boosts for abnormal spikes.

Pseudo-L7 Inference

Classifies HTTP floods, Slowloris, and amplification patterns from flow metadata — enriches attack type and can boost the final score.

≥70
HIGH — Attack confirmed
50–69
MEDIUM — Suspicious
<50
LOW — Filtered out
Automated Mitigation

From Detection to Mitigation

Detection is the foundation of effective DDoS mitigation. Once Smart Detection confirms an attack (score ≥70), CoreDetection automatically triggers mitigation workflows using standard network mechanisms — configured separately and activated only when thresholds are exceeded.

BGP BlackHole / RTBH

Drop all traffic to attacked prefixes upstream — applied only when Smart Detection confirms HIGH confidence (score ≥70).

BGP FlowSpec Filtering

Target specific protocols, ports, or packet characteristics — preserve legitimate traffic during multi-vector attacks.

Telegram, Email & Webhooks

Instant NOC alerting on attack START, UPDATE, and END with full telemetry payload.

Manual Override via API

Announce, withdraw, or inspect BGP routes in real time from dashboard or REST API.

System Modules

Four Modules, One Appliance

CORE 3-Layer Analysis

Smart Detection Engine

CoreDetection ingests NetFlow, IPFIX, and sFlow from your routers. When traffic crosses thresholds, Smart Detection runs three analysis layers — rhythm, fingerprinting, and attack memory — producing a weighted confidence score before any action is taken.

Rhythm Analysis

Analyses traffic cadence and burst patterns from 61-second samples — erratic spikes vs smooth ramps.

Traffic Fingerprinting

Maps source clustering, ASN diversity, and geo dispersion — with trusted-ASN score adjustment.

Attack Memory

Recognises repeat campaigns at 85%+ similarity — instant high-confidence alerts for known patterns.

0–100
Confidence Score
BGP BGP Blackhole & FlowSpec

Automated Mitigation

Built-in GoBGP speaker announces blackhole routes or FlowSpec rules when Smart Detection confirms HIGH confidence (score ≥70). Mitigation uses standard routing mechanisms already present in your network.

BGP Blackhole (RTBH)

Announce /32 host routes with blackhole community to drop attack traffic at the network edge instantly.

BGP FlowSpec (RFC 5575/8955)

Drop, rate-limit, scrubbing redirect (accept/nexthop), RT redirect, or DSCP mark — with TCP flag and fragmentation matching.

IP Blocklist

Ingest threat-intelligence feeds (HTTP/HTTPS/local file) and enforce source-prefix drops via FlowSpec automatically.

Manual Override

Full BGP route management via REST API or dashboard — announce, withdraw, or inspect routes in real time.

Auto
BGP Response
UI Full Operations Visibility

Dashboard & Reports

Built-in operations dashboard with live attack monitoring, Smart Detection score breakdown (rhythm / fingerprint / memory), prefix intelligence, and exportable forensic reports.

Live Attack Monitor

Real-time view of active attacks with severity, bandwidth, and rhythm/fingerprint/memory scores.

Forensic Reports

Full attack history exportable as JSON or CSV. Query by target, severity, date, or attack type.

Alert Center

Telegram, email, and webhook event log with attack lifecycle tracking (START / UPDATE / END).

Built-In
No Add-On
API Programmable Control

REST API & Alerts

Full REST API at /CoreDetection/api/v0 for runtime config, attack reports, and BGP management. Push alerts to Telegram, email, or webhooks — integrate with existing SOC workflows.

RESTful API

Configure thresholds, Smart Detection, BGP, alerts live — no restart required.

Telegram & Email

Direct alerting on attack START, UPDATE, and END events with full telemetry payload.

SIEM Integration

Webhooks and CSV/JSON export for Grafana, ELK, Prometheus, or custom SOC dashboards.

REST
Full Control
Built-In Dashboard

Monitor. Analyze. Report.

Full operations dashboard — no external portal or paid add-on. Live attacks, Smart Detection score breakdowns, BGP control, and exportable forensic reports.

Live Attack Monitor

Real-time view of active attacks, severity, bandwidth, and Smart Detection scores.

Prefix Intelligence

Per-prefix traffic baselines, anomaly trends, and detection history.

BGP Control Panel

Session status, announced routes, blackhole and FlowSpec actions.

Attack Reports

Full forensic history — export as JSON or CSV, query by date, target, or severity.

Smart Detection Insights

Rhythm, fingerprint, and memory score breakdown with attack memory matches.

Alert Center

Telegram, email, and webhook event log with lifecycle tracking.

REST API

Designed for Network Engineers

Full REST API at /CoreDetection/api/v0 on port 9009. Configure Smart Detection, pull reports, and manage BGP without restart.

No proprietary hardware or vendor-specific dependencies required — standard flow export protocols and BGP integration into existing infrastructure.

  • ISPs and telecom operators
  • Data centres and hosting providers
  • Cloud and hybrid infrastructure operators
  • Enterprises with high-value online services
Full API Reference →
localhost:9009/CoreDetection/api/v0
GET /status
GET /attacks/active
GET /reports/attacks
GET /smart-detection/stats
POST /bgp/blackhole
POST /flowspec/apply
GET /blocklist/status
Industries

Built For Your Network

ISPs, data centers, enterprises, and hosting providers — same Smart Detection engine, same dashboard, your infrastructure.

ISPs & Carriers

Monitor customer prefixes from flow telemetry. Auto-blackhole attacked /32s and alert your NOC.

  • Multi-Prefix Monitoring
  • BGP Auto-Mitigation
  • Smart Detection

Data Centers

Detect volumetric floods targeting hosted customers before they saturate your uplinks.

  • Prefix-Level Detection
  • FlowSpec Filtering
  • Dashboard & Reports

Enterprises

AI-powered detection on your own network — full data sovereignty, no cloud dependency.

  • On-Premises Deploy
  • Offline License
  • Attack Reports

Hosting Providers

Protect tenant IP ranges with automated BGP response and per-customer prefix monitoring.

  • Multi-Tenant Prefixes
  • Webhook Alerts
  • Smart Detection

Ready to Deploy?

Install CoreDetection, point your routers, and start detecting with Smart Detection — dashboard and reports included.