| name | implementing-network-traffic-baselining |
| description | Build network traffic baselines from NetFlow/IPFIX data using Python pandas for statistical analysis, z-score anomaly detection, and hourly/daily traffic pattern profiling |
| domain | cybersecurity |
| subdomain | network-security |
| tags | ["netflow","ipfix","traffic-analysis","baselining","anomaly-detection","pandas","network-monitoring"] |
| version | 1.0 |
| author | mahipal |
| license | Apache-2.0 |
| nist_csf | ["PR.IR-01","DE.CM-01","ID.AM-03","PR.DS-02"] |
Implementing Network Traffic Baselining
Overview
Network traffic baselining establishes normal communication patterns by analyzing historical NetFlow/IPFIX data to create statistical profiles of expected behavior. This skill uses Python pandas to compute hourly and daily traffic distributions, per-host byte/packet counts, protocol ratios, and top-N talker profiles. Anomalies are detected using z-score thresholds and IQR (interquartile range) outlier methods, enabling SOC analysts to identify deviations such as data exfiltration spikes, beaconing patterns, and unusual port usage.
When to Use
- When deploying or configuring implementing network traffic baselining capabilities in your environment
- When establishing security controls aligned to compliance requirements
- When building or improving security architecture for this domain
- When conducting security assessments that require this implementation
Common Misconfigurations & Verification
- Z-score assumes normality: network volume is heavy-tailed and diurnal, so a global z-score flags every business-hours peak and misses off-hours exfil. Baseline per hour-of-day/day-of-week and prefer IQR/MAD for skewed data rather than one mean/stddev over all time.
- Baseline window too short or polluted: under ~7 days, or a window that already contains the incident, bakes anomalies into "normal." Confirm the training period is clean and long enough to cover weekly cycles.
- Sampled NetFlow skews volumes: sampled exports (e.g. 1:1000) and active-timeout flow splitting make one long transfer look like many small flows. Normalize for sampling rate and account for
flowDuration before summing bytes/packets.
- Aggregation hides per-host spikes: baselining only total throughput masks a single host's exfil; profile per-source-IP too.
Verification: validate the detector against ground truth before relying on it — inject a synthetic spike (a host sending 10x its baseline bytes in one hour) and confirm it surfaces with a high z-score/IQR flag, and confirm a normal diurnal peak does not. Spot-check that flagged flows are reproducible in the raw NetFlow (matching srcip, bytes, timestamp). A model that fires on every daily peak, or stays silent when you inject a known spike, is mis-tuned — not validating a quiet network.
Prerequisites
- NetFlow v5/v9 or IPFIX flow data exported as CSV or JSON
- Python 3.8+ with pandas and numpy libraries
- Historical flow data (minimum 7 days recommended for baseline)
Steps
