Performing Network Forensics With Wireshark

Performing Network Forensics with Wireshark

When to Use

• When analyzing captured network traffic (PCAP files) from a security incident
• For identifying command-and-control (C2) communications in captured traffic
• When reconstructing data exfiltration activities from packet captures
• During malware analysis to identify network indicators of compromise
• For extracting files, credentials, and artifacts transferred over the network

Detection Gaps & Validation

A PCAP only contains what the sensor actually saw, decrypted to the depth your tooling supports. Account for these blind spots:

• Capture truncation and drops. A snaplen below the MTU (tcpdump -s too small) gives you headers without payloads, so --export-objects silently produces nothing. Run capinfos and check for dropped packets, snaplen, and whether the duration covers the incident window — "no files extracted" is often truncation, not absence. Missing one direction (asymmetric routing/SPAN) breaks stream reassembly.
• Encrypted payloads hide the content, not the metadata. TLS 1.3/QUIC, DoH/DoT, and SSH mean you can't read bodies without keys. Pivot to what stays visible: SNI (often, but ECH/encrypted ClientHello blocks it), JA3/JA3S fingerprints, certificate issuers/validity, connection sizes, and timing. Don't conclude traffic is benign just because you can't decrypt it.
• C2 hides in normal-looking flows. Beaconing with jitter defeats fixed-interval filters; domain fronting and traffic over 443/53 blend in. Validate a suspected beacon by computing inter-arrival variance across the whole capture and correlating the destination with endpoint process/DNS logs, not a single frame.time_delta filter.
• Cross-corroborate and verify, don't trust one heuristic. Reassembled TCP streams can be incomplete if SYN/retransmits are missing; carved files may be partial. Hash every extracted object and check it against threat intel/VirusTotal before calling it malware, and confirm the file actually transferred (response 200 + content-length) versus an aborted request.
• Interpretation false positives. Long/random DNS labels are frequently CDNs, AV telemetry, and analytics — not always tunneling; self-signed certs are common on internal services. NetworkMiner OS/port guesses are heuristic. Verify clock/timezone of the capture host before building a timeline, and treat plaintext "credentials" as candidates until you confirm the service and account.

Prerequisites

• Wireshark or tshark installed for packet analysis
• PCAP/PCAPNG files from network captures (tcpdump, Wireshark, network TAP)
• NetworkMiner for automated artifact extraction
• Sufficient RAM for large capture files (1GB+ PCAPs need 8GB+ RAM)
• Understanding of TCP/IP, HTTP, DNS, TLS protocols
• GeoIP databases for IP geolocation

Workflow

Step 1: Prepare and Validate the Capture File

# Install Wireshark and tshark
sudo apt-get install wireshark tshark

# Verify the PCAP file
capinfos /cases/case-2024-001/network/capture.pcap

# Output includes: file type, packet count, capture duration, data size
# Example output:
# File name:           capture.pcap
# File type:           Wireshark/tcpdump/... - pcap
# Number of packets:   1,245,678
# File size:           856 MB
# Data size:           823 MB
# Capture duration:    3600.123456 seconds
# First packet time:   2024-01-15 14:00:00.000000
# Last packet time:    2024-01-15 15:00:00.123456

# Hash the PCAP for integrity
sha256sum /cases/case-2024-001/network/capture.pcap \
   > /cases/case-2024-001/network/pcap_hash.txt

# Get a protocol hierarchy statistics overview
tshark -r /cases/case-2024-001/network/capture.pcap -q -z io,phs

Step 2: Filter and Identify Suspicious Traffic

# Extract conversation statistics
tshark -r /cases/case-2024-001/network/capture.pcap -q -z conv,tcp

# Find top talkers by bytes transferred
tshark -r /cases/case-2024-001/network/capture.pcap -q -z endpoints,ip \
   | sort -t$'\t' -k3 -rn | head -20

# Filter for DNS queries (potential C2 or exfiltration)
tshark -r /cases/case-2024-001/network/capture.pcap \
   -Y "dns.qr == 0" \
   -T fields -e frame.time -e ip.src -e dns.qry.name \
   > /cases/case-2024-001/analysis/dns_queries.txt

# Find DNS queries to unusual TLDs or long domain names (DNS tunneling)
tshark -r /cases/case-2024-001/network/capture.pcap \
   -Y "dns.qr == 0 && dns.qry.name matches \"[a-z0-9]{30,}\"" \
   -T fields -e frame.time -e ip.src -e dns.qry.name \
   > /cases/case-2024-001/analysis/suspicious_dns.txt

# Filter HTTP traffic
tshark -r /cases/case-2024-001/network/capture.pcap \
   -Y "http.request" \
   -T fields -e frame.time -e ip.src -e ip.dst -e http.request.method \
   -e http.host -e http.request.uri -e http.user_agent \
   > /cases/case-2024-001/analysis/http_requests.txt

# Find connections to known malicious ports
tshark -r /cases/case-2024-001/network/capture.pcap \
   -Y "tcp.dstport == 4444 || tcp.dstport == 8080 || tcp.dstport == 1337 || tcp.dstport == 6667" \
   -T fields -e frame.time -e ip.src -e ip.dst -e tcp.dstport \
   > /cases/case-2024-001/analysis/suspicious_ports.txt

# Detect beaconing patterns (regular interval connections)
tshark -r /cases/case-2024-001/network/capture.pcap \
   -Y "ip.dst == 185.0.0.1" \
   -T fields -e frame.time_epoch \
   > /tmp/beacon_times.txt

Step 3: Extract Files and Objects from Traffic

# Export HTTP objects (files transferred over HTTP)
tshark -r /cases/case-2024-001/network/capture.pcap \
   --export-objects http,/cases/case-2024-001/analysis/http_objects/

# Export SMB objects
tshark -r /cases/case-2024-001/network/capture.pcap \
   --export-objects smb,/cases/case-2024-001/analysis/smb_objects/

# Export DICOM objects (medical imaging)
tshark -r /cases/case-2024-001/network/capture.pcap \
   --export-objects dicom,/cases/case-2024-001/analysis/dicom_objects/

# Export FTP data transfers
tshark -r /cases/case-2024-001/network/capture.pcap \
   -Y "ftp-data" \
   -T fields -e ftp-data.data \
   --export-objects ftp-data,/cases/case-2024-001/analysis/ftp_objects/

# Hash all extracted objects
find /cases/case-2024-001/analysis/http_objects/ -type f -exec sha256sum {} \; \
   > /cases/case-2024-001/analysis/extracted_file_hashes.txt

# Check extracted file hashes against VirusTotal
while read hash filepath; do
   echo "Checking $filepath ($hash)"
   curl -s "https://www.virustotal.com/api/v3/files/$hash" \
      -H "x-apikey: YOUR_API_KEY" | python3 -c "
import json,sys
data=json.load(sys.stdin)
if 'data' in data:
   stats=data['data']['attributes']['last_analysis_stats']
   print(f'  Malicious: {stats[\"malicious\"]}, Undetected: {stats[\"undetected\"]}')
else:
   print('  Not found on VT')
"
done < /cases/case-2024-001/analysis/extracted_file_hashes.txt

Step 4: Reconstruct TCP Streams and Sessions

# Follow a specific TCP stream (stream index 42)
tshark -r /cases/case-2024-001/network/capture.pcap \
   -q -z "follow,tcp,ascii,42" \
   > /cases/case-2024-001/analysis/stream_42.txt

# Extract all HTTP request-response pairs for a suspicious host
tshark -r /cases/case-2024-001/network/capture.pcap \
   -Y "http && ip.addr == 185.0.0.1" \
   -T fields -e frame.time -e http.request.method -e http.host \
   -e http.request.uri -e http.response.code -e http.content_length \
   > /cases/case-2024-001/analysis/suspicious_http.txt

# Extract TLS/SSL certificate information
tshark -r /cases/case-2024-001/network/capture.pcap \
   -Y "tls.handshake.type == 11" \
   -T fields -e ip.dst -e tls.handshake.certificate \
   > /cases/case-2024-001/analysis/tls_certs.txt

# Extract TLS SNI (Server Name Indication) values
tshark -r /cases/case-2024-001/network/capture.pcap \
   -Y "tls.handshake.extensions_server_name" \
   -T fields -e frame.time -e ip.src -e ip.dst \
   -e tls.handshake.extensions_server_name \
   > /cases/case-2024-001/analysis/tls_sni.txt

# Extract credentials from unencrypted protocols
tshark -r /cases/case-2024-001/network/capture.pcap \
   -Y "ftp.request.command == \"USER\" || ftp.request.command == \"PASS\"" \
   -T fields -e frame.time -e ip.src -e ftp.request.command -e ftp.request.arg

tshark -r /cases/case-2024-001/network/capture.pcap \
   -Y "http.authorization" \
   -T fields -e frame.time -e ip.src -e http.host -e http.authorization

Step 5: Use NetworkMiner for Automated Analysis

# Install NetworkMiner (Mono required on Linux)
sudo apt-get install mono-complete
wget https://www.netresec.com/?download=NetworkMiner -O NetworkMiner.zip
unzip NetworkMiner.zip -d /opt/NetworkMiner/

# Run NetworkMiner
mono /opt/NetworkMiner/NetworkMiner.exe /cases/case-2024-001/network/capture.pcap

# NetworkMiner automatically extracts:
# - Host inventory (OS fingerprinting, open ports)
# - Files transferred over HTTP, FTP, SMB, TFTP
# - Images from web traffic
# - Credentials (plaintext and NTLM hashes)
# - DNS records
# - Session parameters
# - Anomalies and alerts

Step 6: Generate Network Forensics Report

# Compile findings
cat << 'EOF' > /cases/case-2024-001/analysis/network_forensics_report.txt
NETWORK FORENSICS ANALYSIS REPORT
===================================
Case: 2024-001
Capture File: capture.pcap (856 MB, 1,245,678 packets)
Capture Period: 2024-01-15 14:00 to 15:00 UTC
Analyst: [Examiner Name]

TRAFFIC OVERVIEW:
  Total packets: 1,245,678
  Unique source IPs: 45
  Unique destination IPs: 234
  Protocols: TCP (78%), UDP (18%), ICMP (2%), Other (2%)

C2 COMMUNICATION:
  Destination: 185.0.0.1:443
  Beaconing interval: ~60 seconds
  Total connections: 58
  Data transferred: 4.2 MB outbound, 12.3 MB inbound
  TLS SNI: update-service.malware-c2.com

EXFILTRATION:
  Method: HTTPS POST to 185.0.0.1
  Volume: 4.2 MB over 45 minutes
  Files: 3 ZIP archives extracted from HTTP objects

DNS TUNNELING:
  Suspicious queries to: data.evil-dns.com
  Average subdomain length: 45 characters
  Query count: 1,234 (normal baseline: 50)
EOF

Key Concepts

Tools & Systems

Common Scenarios

Scenario 1: Malware C2 Communication Analysis Load PCAP in Wireshark, identify beaconing patterns to external IPs, examine TLS certificates for self-signed or unusual issuers, extract HTTP POST data containing encoded commands, correlate C2 IPs with threat intelligence feeds.

Scenario 2: Data Exfiltration Detection Analyze traffic statistics for unusually large outbound transfers, examine DNS query lengths for DNS tunneling indicators, track FTP and HTTP file uploads to external servers, reconstruct exfiltrated files from packet data.

Scenario 3: Lateral Movement in Enterprise Network Filter for SMB, RDP, WMI, and PSExec traffic between internal hosts, identify credential usage patterns across multiple systems, trace the propagation path of the attacker through the network, correlate with Windows Event Log authentication events.

Scenario 4: Web Application Attack Reconstruction Filter HTTP traffic to the web server, identify SQL injection, XSS, and directory traversal attempts, follow the TCP stream of the successful exploit, extract uploaded webshells or payloads, document the attack chain for the incident report.

Output Format

Network Forensics Summary:
  Capture: capture.pcap
  Duration: 1 hour (14:00-15:00 UTC, 2024-01-15)
  Packets: 1,245,678 | Size: 856 MB

  Top Suspicious Connections:
    192.168.1.50 -> 185.0.0.1:443   (C2, 58 connections, 4.2MB out)
    192.168.1.50 -> 10.0.0.25:445   (SMB lateral movement)
    192.168.1.50 -> 10.0.0.30:3389  (RDP lateral movement)

  Extracted Artifacts:
    Files:        23 (3 malicious per VT)
    Credentials:  2 plaintext FTP logins
    DNS Queries:  1,234 suspicious (possible tunneling)
    TLS Certs:    5 self-signed certificates

  IOCs Identified:
    IPs:     185.0.0.1, 203.0.113.50
    Domains: update-service.malware-c2.com, data.evil-dns.com
    Hashes:  3 file hashes flagged as malware