Portfolio Reference: AM-SEC-PORT-001  |  Classification: Public — Engineering Portfolio  |  Version: 1.0
Security Architecture · Detection Engineering · SOC Operations

Alec Sanchez

Cybersecurity Engineer — Early Career

This portfolio presents original security architecture research, enterprise-grade homelab implementation, and structured detection engineering work. The primary contribution is the Orthogonal Structure Invariant Model (OSIM) — a five-zone network segmentation framework with machine-verifiable security invariants, implemented and validated in a production-equivalent lab environment.

5×
VLAN Security Zones
50+
Firewall Rules
10
Security Invariants

Orthogonal Structure Invariant Model (OSIM)

A multi-zone network segmentation framework engineered for blast radius containment, least-privilege network access, and machine-verifiable security properties.

1.1 — Framework Purpose

OSIM addresses a structural failure mode common in flat and semi-segmented networks: implicit trust derived from network position. The model partitions all hosts into five discrete trust zones, enforces default-deny at every inter-zone boundary, and defines ten binary machine-verifiable invariants whose violation constitutes a critical governance event. The design objective is blast radius containment — ensuring that compromise of any single zone cannot produce lateral movement to higher-trust zones through network-layer paths alone.

1.2 — Zone Definitions and Trust Boundaries

Zone VLAN Subnet Trust Tier Boundary Function
MGMT-A — Admin Workstations 10 10.10.10.0/24 Tier 0-User Jump hosts and administrative endpoints. Initiates management sessions to MGMT-B on explicitly scoped ports only.
MGMT-B — Infrastructure 11 10.10.11.0/24 Tier 0-Infra Proxmox hypervisor, OPNsense management IP, Proxmox Backup Server. No workstations. Isolated from MGMT-A by port-scoped rules.
LAB — Compromise Domain 20 10.10.20.0/24 Tier 2 Active red/blue operations, attack simulation targets. Default-deny terminator L-99 prevents all lateral movement to higher-trust zones.
DMZ — Exposed Services 30 10.10.30.0/24 Tier 1-Ext Internet-facing reverse proxy and application servers. Zero inward connectivity to any internal zone except Wazuh agent ports.
SECURITY — SIEM Plane 40 10.10.40.0/24 Tier 0-Det Wazuh Manager, detection tooling. Accepts telemetry from all zones. Initiates no management sessions to any zone.

1.3 — Design Principles

P-01

No Implicit Trust

Network position confers zero trust. Every inter-zone communication path requires an explicit, port-scoped allow rule with documented justification. No ANY-ANY rules exist on any interface.

P-02

Default-Deny Terminator Model

Each VLAN interface terminates with an explicit logged block-all rule. The absence of a matching allow rule produces a deny-and-log event — not a silent pass. This eliminates over-permissive rules introduced through operational convenience.

P-03

Administrative Tier Decoupling

MGMT-A (workstations) and MGMT-B (infrastructure) are separated into distinct VLANs. A compromised administrator workstation does not confer access to hypervisor management interfaces. Lateral movement requires a separately authenticated session, independently logged.

P-04

Detection Plane Isolation

The SIEM plane (VLAN 40) initiates no management connections. Active response is agent-based only. No network path exists from any monitored zone to the Wazuh API — the detection plane cannot be silenced by compromise of a monitored zone.

P-05

Configuration Integrity as Code

The OPNsense running configuration is exported, SHA-256 hashed, and committed to version control after every change. Daily automated hash comparison detects unauthorized drift from the committed baseline.

1.4 — Ten Machine-Verifiable Security Invariants

Each invariant is binary — TRUE or VIOLATED. Violation constitutes a critical governance event requiring immediate investigation. Invariant verification is automated via configuration XML parser.

INV-01
No ANY-ANY pass rule exists on any interface
INV-02
MGMT-A → MGMT-B: only TCP/8006, 22, 443 permitted
INV-03
No DMZ pass rule to any internal zone except Wazuh agent ports
INV-04
Each interface terminates with explicit logged block-all rule
INV-05
All DNS queries route through internal Unbound resolver only
INV-06
No LAB zone pass rule to MGMT-A, MGMT-B, or SECURITY
INV-07
WAN interface: no pass rules to MGMT-A, MGMT-B, or SECURITY
INV-08
SECURITY zone initiates no outbound management sessions to any zone
INV-09
OPNsense configuration XML hash matches committed Git baseline
INV-10
No SSH path exists from SECURITY zone to any infrastructure host

Security Architecture Research

Original technical documentation produced to enterprise architecture standards, aligned to NIST SP 800-53, Zero Trust principles, and MITRE ATT&CK framework mappings.

OSIM-FW-ARCH-001 Policy Version: v1.1.1 Platform: OPNsense 24.x / Proxmox VE Enforcement: Default-Deny

Firewall Segmentation Architecture: Enterprise Network Zoning, Trust Boundary Enforcement, and Detection Integration

Describes a multi-zone network segmentation architecture implemented on OPNsense 24.x (FreeBSD 14) with stateful packet filtering (pf) and inline Suricata IDS/IPS. Documents north–south traffic enforcement, east–west segmentation controls, Suricata IDS/IPS deployment with VLAN de-encapsulation verification, Wazuh SIEM integration, governance framework, and structured red team validation.

Primary Research NIST SP 800-53 Zero Trust MITRE ATT&CK OPNsense / pf Suricata IDS/IPS Wazuh SIEM Red Team Validation

Risk Reduction Outcomes

Risk Scenario Inherent Controls Applied Residual
Lab VM pivots to hypervisor (Proxmox/OPNsense) HIGH L-99 terminator; VLAN 11 isolation; intra-infra logging; Wazuh deny alerting LOW
Compromised DMZ host pivots to internal zones HIGH D-99 terminator; no pass rules to any internal zone except Wazuh agent ports LOW
Admin workstation compromise escalates to infrastructure HIGH MA-08 explicit block; port-scoped allow (8006/22/443 only); separate VLANs 10 and 11 MED*
DNS tunneling / C2 via external resolver MED DNS-BLOCK floating rule; Suricata ET DNS signatures; Unbound as sole permitted resolver LOW
HTTPS C2 egress from LAB or DMZ (port 443) MED Suricata JA3 fingerprint detection; ET MALWARE ruleset; TLS metadata inspection MED*
SIEM disruption from compromise domain HIGH Ports 1514/1515 only; S-05 blocks all other inbound to Wazuh; L-99/D-99 terminators LOW
Unauthorized internet access to management interfaces HIGH W-IN-02 WAN default-deny; no WAN pass rules to MGMT-A, MGMT-B, or SECURITY LOW
Accepted Residuals * MGMT-A → MGMT-B: Port-restricted access is formally accepted at current architecture maturity. Mitigation path: Privileged Access Workstation (PAW) with MFA enforcement on Proxmox and OPNsense.
* HTTPS C2 Egress: Outbound TCP/443 cannot be blocked without degrading operational capability. Mitigation path: TLS inspection proxy (Squid with SSL bump).

Homelab Architecture

A production-equivalent security operations environment built on Proxmox VE with OPNsense firewall, five-zone VLAN segmentation, and full telemetry integration.

Hypervisor Platform

PlatformProxmox VE
BackupProxmox PBS
SnapshotsPre-change gating
Cluster commsVLAN 11 (MGMT-B)
Mgmt interfaceTCP/8006 only

Firewall / Network

PlatformOPNsense 24.x
OSFreeBSD 14
Filter enginepf (stateful)
Zones5 VLANs
DNS resolverUnbound (local only)
NATPure NAT

Security Stack

IDS/IPSSuricata (trunk)
RulesetET Open
SIEMWazuh Manager
Agent coverageAll 5 zones
Log formateve.json + syslog/514
Config mgmtGit + SHA-256

Suricata IDS/IPS Deployment

Suricata is deployed on the trunk interface (vtnet1), upstream of sub-interface VLAN processing. This positions Suricata to inspect all inter-zone traffic on a single capture point, regardless of source and destination VLAN. VLAN de-encapsulation is explicitly configured in suricata.yaml (vlan: use-for-tracking: true). Without this, Suricata attempts signature matching against raw 802.1Q-tagged frames, resulting in incorrect IP header offset calculations and silently absent content-based detection.

IPS Promotion Gate Promotion from IDS mode to inline IPS drop mode is gated on five criteria: de-encapsulation verification, a 14-day false positive baseline below 5 alerts/day, full suppression list documentation, confirmed Wazuh parsing of eve.json, and a Proxmox VM snapshot of the OPNsense host.

East–West Segmentation Controls

Boundary Permitted Traffic Enforcement ATT&CK Techniques Mitigated
MGMT-A → MGMT-B TCP/8006, TCP/22, TCP/443 Rules MA-04–MA-07; MA-08 blocks all other T1611 Escape to Host, T1078 Valid Accounts, T1098 Account Manipulation
LAB → MGMT-A/B None Default-deny terminator L-99 T1021.004 SSH, T1021.001 RDP, T1563 Session Hijacking, T1078 Valid Accounts
LAB → SECURITY TCP/1514–1515 to HOST_WAZUH only Wazuh agent permit L-01/L-02 ahead of L-99 T1562.001 Disable Tools, T1070 Indicator Removal, T1499 DoS (SIEM)
DMZ → Internal (all) TCP/1514–1515 to HOST_WAZUH only Default-deny terminator D-99 T1021 Remote Services, T1090.003 Multi-hop Proxy, T1190 Exploit Public App
SECURITY → MGMT-B None Explicit block rules S-08/S-09 T1219 Remote Access Software (detection-plane pivot)
Intra-DMZ (lateral) Proxy → App backend TCP/80, 443 only Rule D-05 blocks all other intra-DMZ traffic T1570 Lateral Tool Transfer

Detection Engineering & Validation

Structured adversary simulation scenarios with Wazuh-verified detection evidence. Telemetry correlated across network deny events, IDS alerts, and host-based agent data.

Red Team Validation Scenarios

RT-01

Hypervisor Pivot Prevention

Validates that a compromised LAB host cannot reach Proxmox management (TCP/8006) or OPNsense management interfaces through any available network path.

Evidence: Wazuh deny-event alert from L-99 terminator
RT-02

DNS Exfiltration Containment

Validates DNS tunneling is blocked at the network layer for all internal zones and detected by Suricata ET DNS high-entropy subdomain signatures.

Evidence: DNS-BLOCK floating rule log + Suricata ET alert
RT-03

Reverse Shell Containment

Validates outbound C2 channel detection via Suricata JA3 fingerprinting and Emerging Threats MALWARE ruleset, with correlated Wazuh host-agent telemetry.

Evidence: Suricata JA3 alert + Wazuh correlated event

MITRE ATT&CK Coverage — Addressed Techniques

T1021.001 RDP T1021.004 SSH T1048.003 DNS Exfil T1071.004 DNS C2 T1078 Valid Accounts T1090.003 Multi-hop Proxy T1190 Exploit Public App T1219 Remote Access T1499 DoS (SIEM) T1562.001 Disable Tools T1563 Session Hijacking T1570 Lateral Tool Transfer T1611 Escape to Host T1070 Indicator Removal T1098 Account Manipulation
Highlighted
Mitigated at network layer
Unhighlighted
Detected via Suricata/Wazuh

Telemetry Coverage by Zone

MGMT-A (Admin)
85%
MGMT-B (Infrastructure)
95%
LAB (Compromise Domain)
90%
DMZ (Exposed Services)
95%
SECURITY (SIEM Plane)
80%

Wazuh SIEM Integration

Detection Pipeline OPNsense ships firewall deny-event logs to Wazuh via syslog UDP/514 in JSON format. Suricata ships eve.json alerts via the Wazuh agent on the OPNsense host. Host-based Wazuh agents are deployed on endpoints in all five zones, providing correlated visibility across network deny events, IDS alerts, and host telemetry in a single SIEM view. The VLAN field in eve.json enables zone-contextual severity — the same Suricata signature from VLAN 20 (LAB) and VLAN 30 (DMZ) produces differentiated alert classifications.

Security Operations Analysis

Structured SOC alert triage, investigation workflow, and threat detection analysis performed using Wazuh SIEM and Suricata IDS telemetry from the homelab environment.

SOC Investigation Workflow

SOC-01

Alert Intake

Alerts originate from Suricata IDS signatures, firewall deny events, and Wazuh host agents. Events are ingested into Wazuh Manager and correlated across network and host telemetry to determine severity and potential threat classification.

SOC-02

Triage & Contextual Analysis

Each alert is evaluated against network zone context (OSIM trust tiers), known benign activity, and MITRE ATT&CK mappings. Alerts are categorized into False Positive, Suspicious Activity, or Confirmed Malicious Behavior.

SOC-03

Investigation & Correlation

Network telemetry (Suricata eve.json), firewall logs, and host agent events are correlated to determine attacker behavior, persistence attempts, or lateral movement indicators.

SOC-04

Containment & Documentation

Confirmed malicious activity results in containment actions such as firewall rule updates, host isolation, or IDS rule promotion. All incidents are documented with timeline analysis and ATT&CK technique mapping.

Example Alert Investigation

Alert Source Detection Zone Analysis Outcome
Suricata ET DNS High Entropy Query LAB Potential DNS tunneling attempt detected. Query analyzed for entropy and repeated subdomain patterns. Blocked + Logged
Firewall L-99 Deny Event LAB → MGMT-B Attempted SSH connection from compromised lab host to infrastructure network. Blocked by segmentation
Suricata JA3 TLS Fingerprint Match DMZ TLS fingerprint matched known malware C2 signature. Alert escalated

SOC Detection Metrics

False Positive Rate
~12%
Average Alert Triage Time
~5 min
Detection Coverage (all zones)
~90%
Operational Objective The SOC analysis environment enables continuous validation of OSIM network segmentation controls while developing practical alert triage and threat investigation workflows consistent with enterprise security operations centers.

CyberDefenders Investigation Labs

Practical threat investigation exercises completed through the CyberDefenders DFIR training platform. Each lab simulates real-world SOC incident scenarios requiring log analysis, network forensics, and attacker behavior identification.

Lab Case Study — Malicious DNS Activity

Lab Focus Artifacts Analyzed Key Finding Techniques
DNS Exfiltration Investigation Network Forensics PCAP + DNS Logs High-entropy DNS queries indicated potential data exfiltration. MITRE ATT&CK T1048
Investigation Summary Traffic analysis identified repeated DNS requests containing encoded subdomain payloads. Entropy analysis and query pattern review confirmed the activity as simulated DNS tunneling behavior used for covert data exfiltration.

Diagram Roadmap

Planned visual documentation to supplement framework and detection sections.

D-01

Five-Zone Trust Architecture (Mermaid / Draw.io)

Network topology with all five VLANs, trust tier color-coding, inter-zone allow/deny paths, and Wazuh telemetry flows. Embed as interactive Mermaid on the OSIM page.

D-02

Suricata Deployment Diagram

Trunk interface positioning, VLAN de-encapsulation flow, and eve.json → Wazuh pipeline. Illustrates single-capture-point coverage across all inter-zone traffic.

D-03

MITRE ATT&CK Navigator Heatmap

Export ATT&CK Navigator JSON showing techniques mitigated (network block) vs. detected (Suricata/Wazuh alert). Embed as screenshot with link to JSON source.

D-04

North–South Traffic Flow Diagram

WAN → DMZ NAT path, per-zone egress NAT, DNS enforcement floating rule position. Include anti-spoofing rule (W-IN-01) and Pure NAT reflection annotation.

D-05

SIEM Detection Pipeline

Data flow: OPNsense pf deny log → syslog UDP/514 → Wazuh Manager. Suricata eve.json → Wazuh agent on OPNsense host → Manager. Host agents all zones → correlated dashboard.

Contact Information

Available for SOC Analyst, Security Operations, Junior Security Engineer, and Cybersecurity Analyst roles.

For Hiring Managers — Quick Reference

Primary frameworkOSIM v1.1.1
ATT&CK techniques addressed15
Security invariants10 (binary / verifiable)
NIST SP 800-53 alignmentSC-7, AU-6, AC-4, SI-4
Red team validation scenariosRT-01, RT-02, RT-03
Certification in progressPJPT (TCM Security)