Final-Year Telecommunications Engineering Student Securing connected vehicles. Detecting threats. Automating defense.
d0ok@github:~$ whoami
I work at the intersection of automotive cybersecurity and security operations. My projects are built to be reproducible: I implement the mechanism or the attack, test it, then document the limitations honestly.
- Automotive security: SecOC-style message authentication, CAN intrusion detection, secure OTA, TARA (ISO/SAE 21434 style), AUTOSAR-style software components
- SOC / Blue Team: Wazuh detection engineering, SOAR automation and threat-intel enrichment, MITRE ATT&CK coverage analysis, ML anomaly detection
- Engineering: C, Python, Linux, Docker, Terraform, Ansible, CI/CD
| Project | What it does | Stack |
|---|---|---|
| secure-ecu-simulator | AUTOSAR-inspired secure ECU in C: HMAC-SHA-256 message authentication, freshness counters with replay protection, persistent counter across restarts. Unit tests, fuzzing under ASan/UBSan, live SocketCAN integration test, CI and a threat model. | C, SocketCAN, CAN FD, OpenSSL |
| can-ids | Unsupervised CAN intrusion detection on the HCRL Car-Hacking dataset. Hybrid ID-whitelist + Isolation Forest reaches F1 0.90–0.96 per attack type; an LSTM baseline is documented as a negative result. Includes a threat model and security report. | Python, scikit-learn, TensorFlow |
| secure-ota-pipeline | Secure vehicle OTA CI/CD demo: CycloneDX SBOM, Grype vulnerability gate, GPG signing in a protected environment, mock backend, simulated ECU client with anti-rollback and A/B slots, plus a UN R156 gap analysis. | GitHub Actions, GPG, Grype, Flask, Python |
| can-security-lab | Replay, injection and DoS attacks on a virtual CAN bus with a detector for each, and an ISO/SAE 21434-style TARA. Honest about simulation limits: the flood test showed monitoring tools losing frames. | Python, SocketCAN, can-utils |
| can-sniffer-analyzer | Simulates 6 ECUs on a virtual CAN bus (vcan0), captures the traffic and produces per-ID statistics (rate, period, jitter, varying bytes) with frequency, timeline and jitter plots. The traffic baseline that can-security-lab builds on. |
Python, SocketCAN, can-utils |
| autosar-intelligent-lighting | Intelligent lighting function as five AUTOSAR-style SWCs, validated MIL and SIL (generated C) against a Python golden model across 16 scenarios, with ROS 2 replay and a Unity 3D demo. | MATLAB/Simulink, Embedded Coder, Python, ROS 2, Unity |
| Project | What it does | Stack |
|---|---|---|
| soc-coverage-analyzer | Maps Wazuh rules and declared log sources to MITRE ATT&CK, classifies coverage and telemetry gaps deterministically, and exports ATT&CK Navigator layers and reports. An optional local LLM (Ollama) only drafts recommendations, and its output is validated and flagged for human review. | Python, Wazuh, MITRE ATT&CK, Pydantic, Ollama, Streamlit |
| soc-as-code | Lightweight SOC pipeline deployed as code: Wazuh detects, n8n enriches alerts in parallel with VirusTotal, AbuseIPDB and AlienVault OTX, scores them, and opens DFIR-IRIS cases automatically. Active Response blocks SSH brute force. Docker Compose and Ansible (with Vault) deployment, validated component by component. | Wazuh, n8n, DFIR-IRIS, Docker, Ansible |
| siem-mobile-kpi | Real-time 4G/5G KPI supervision (RSRP, RSRQ, RSCP, Ec/N0) with custom Wazuh rules and Isolation Forest anomaly detection, on 18k+ field measurements. | Wazuh, Elasticsearch, Kibana, Filebeat, Python, Docker Compose |
| Project | What it does | Stack |
|---|---|---|
| vpn-devsecops | DevSecOps pipeline for an IPSec / TLS VPN: Terraform and Ansible deployment, Checkov and OPA policy gates, and automated attack tests (IKE brute force, downgrade, DoS) with ACL validation. | Terraform, Ansible, strongSwan, OpenVPN, Checkov, OPA, Scapy, GitHub Actions |
Automotive: CAN / CAN FD · SocketCAN · SecOC concepts · UN R156 · ISO/SAE 21434 (TARA) · AUTOSAR-style SWCs · Simulink MIL/SIL
SOC: Wazuh · ELK · n8n (SOAR) · DFIR-IRIS · VirusTotal / AbuseIPDB / OTX enrichment · MITRE ATT&CK · detection engineering · Isolation Forest
Supply chain: CycloneDX SBOM · Grype · GPG signing · anti-rollback
DevSecOps: Terraform · Ansible · Checkov · OPA · strongSwan · GitHub Actions
- AUTOSAR Classic mapping and ARXML generation for the lighting project
- Counter resynchronization and key rotation for the secure ECU simulator
- Live ROS 2 link between the lighting results and the Unity viewer