Hunting For Command And Control Beaconing

Hunting for Command and Control Beaconing

When to Use

• When proactively hunting for compromised systems in the network
• After threat intel indicates C2 frameworks targeting your industry
• When investigating periodic outbound connections to suspicious domains
• During incident response to identify active C2 channels
• When DNS query logs show unusual patterns to specific domains

Detection Gaps & Validation

• Jitter and long sleep break naive frequency rules: interval + 0–50% jitter (Cobalt Strike) and multi-hour sleeps drop the connection count and inflate CV — score on bytes-out consistency and use autocorrelation over a 7–14 day window, not a single-day CV < 0.2.
• Encrypted/allowlisted channels: TLS (T1573), domain fronting, and C2 over Slack/Discord/Telegram/Graph API ride trusted FQDNs — destination reputation alone misses them; pivot to JA3/JA4 (use Zeek ssl.log/Suricata, since Sysmon EID 3 lacks it) and timing.
• DNS C2 (T1071.004): TXT/CNAME tunneling needs DNS query logs (Sysmon EID 22 or resolver logs) plus entropy/volume analysis — invisible if DNS logging is off or only NXDOMAIN is retained.
• Log fidelity: sampled NetFlow or aggregated proxy logs destroy inter-arrival timing — confirm per-connection records.
• Validate: run Atomic Red Team T1071.001/T1071.004 (or a Sliver test profile) with a known interval and confirm the beacon search and DNS-entropy search fire, then map source IP to host via Sysmon EID 1/3.
• Tune FPs: Windows Update, AV, NTP, telemetry, and CDN health checks are periodic by design — allowlist by destination + process, never by interval alone.

Prerequisites

• Network proxy/firewall logs with full URL and timing data
• DNS query logs (passive DNS, DNS server logs, or Sysmon Event ID 22)
• Zeek/Bro network connection logs or NetFlow data
• SIEM with statistical analysis capabilities (Splunk, Elastic)
• Threat intelligence feeds for domain/IP reputation

Workflow

1. Identify Beaconing Characteristics: Define what constitutes beaconing (regular intervals, small payload sizes, consistent destinations, jitter patterns).
2. Collect Network Telemetry: Aggregate proxy logs, DNS queries, and connection metadata for analysis.
3. Apply Frequency Analysis: Identify connections with regular intervals using statistical methods (standard deviation, coefficient of variation).
4. Filter Known-Good Traffic: Exclude legitimate periodic traffic (Windows Update, AV updates, heartbeat services, NTP).
5. Analyze Domain/IP Reputation: Check identified beaconing destinations against threat intel, WHOIS data, and certificate transparency logs.
6. Investigate Endpoint Context: Correlate beaconing activity with process creation, user context, and file system changes on source endpoints.
7. Confirm and Respond: Validate C2 activity, block communication, and initiate incident response.

Key Concepts

Tools & Systems

Common Scenarios

1. Cobalt Strike Beacon: HTTP/HTTPS beaconing with configurable sleep time and jitter to malleable C2 profiles.
2. DNS Tunneling C2: Data exfiltration and command receipt via encoded DNS TXT/CNAME queries to attacker-controlled domains.
3. Sliver C2 over HTTPS: Modern C2 framework using HTTPS with configurable beacon intervals and domain fronting.
4. DGA-based C2: Malware generating random domains daily, with adversary registering upcoming domains for C2.
5. Legitimate Service Abuse: C2 over legitimate cloud services (Azure, AWS, Slack, Discord, Telegram).

Output Format

Hunt ID: TH-C2-[DATE]-[SEQ]
Source IP: [Internal IP]
Source Host: [Hostname]
Destination: [Domain/IP]
Protocol: [HTTP/HTTPS/DNS/Custom]
Beacon Interval: [Average seconds]
Jitter: [Percentage]
Connection Count: [Total connections]
Data Volume: [Bytes sent/received]
First Seen: [Timestamp]
Last Seen: [Timestamp]
Domain Age: [Days]
TI Match: [Yes/No - source]
Risk Level: [Critical/High/Medium/Low]