Adversary Hardening Tiers

The arms race has structure. Map it.


Purpose

ASTRA techniques do not apply uniformly. A $200 toy drone and a military UAS with encrypted links and GPS authentication present fundamentally different attack surfaces. Hardening tiers map target sophistication to technique viability.

Assessed during Phase B (Identification) and Phase F (pre-engagement intelligence). Feeds Phase D (Decision), filtering technique categories to only those viable against the assessed level.


Tier 0 — COTS Unmodified

Factory-default configuration. No security hardening applied.

Characteristic Typical
Command protocol MAVLink v1/v2, unsigned, unencrypted
GPS Civil L1, no authentication
Firmware Factory version, publicly documented
Telemetry Unencrypted, broadcast in the clear
Failsafe logic Published, deterministic, exploitable

Viable categories: All five. Full ASTRA toolkit applies.

Prevalence: The majority of commercially available platforms, including Chinese COTS drones used for ISR. This is the current baseline threat.


Tier 1 — Protocol-Hardened

Communication protocols secured. Command authentication and/or encryption enabled. Everything else remains Tier 0.

Characteristic Typical
Command protocol MAVLink v2 with signing, or encrypted proprietary
GPS Civil L1, no authentication (unchanged)
Telemetry Encrypted or signed

Viable categories:

Category Status
Specter ✓ Encrypted traffic still detectable via RF fingerprinting
Mirage ✓ GPS spoofing unaffected by protocol hardening
Fracture ✓ Decision logic unaffected
Override ✗ Requires defeating signing/encryption
Sever ✓ Resource exhaustion and comms denial unaffected

Protocol hardening defeats Override but leaves everything else intact. A Tier 1 target is marginally harder than Tier 0.


Tier 2 — Sensor-Hardened

Sensor inputs hardened. GPS authentication, multi-sensor fusion, anti-spoofing firmware, or redundant navigation. Protocol hardening also present.

Characteristic Typical
Command protocol Encrypted and authenticated
GPS Multi-constellation with authentication, or GPS-denied navigation
Firmware Hardened, anti-tamper protections may be present
Navigation Multi-modal (GPS + visual + LIDAR + INS)

Viable categories:

Category Status
Specter ✓ Detection works; classification harder
Mirage ⚠ GPS spoofing defeated by authentication. Multi-modal requires multi-sensor spoofing
Fracture ⚠ Harder with adaptive logic but decision complexity creates new attack surface
Override ✗ Protocol hardening defeats injection
Sever ✓ Physics-based attacks unaffected

Significant step up. Fracture and Sever become primary viable active categories.


Tier 3 — Full-Stack Hardened

Every layer hardened. Military-grade or purpose-built.

Characteristic Typical
Command protocol Military encryption, hardware security modules
GPS Military M-code or GPS-denied capable
Firmware Signed, encrypted, anti-tamper, secure boot
Telemetry LPI waveforms
Failsafe logic Adaptive, non-deterministic
Physical Anti-tamper enclosures, potted electronics

Viable categories:

Category Status
Specter ⚠ RF detection viable but classification severely limited
Mirage ✗ Without novel research
Fracture ⚠ Requires novel research
Override ✗ Cryptographic protection defeats known techniques
Sever ⚠ Resource exhaustion works (physics). Anti-jam degrades comms denial

Phase D must honestly report “no viable countermeasure with current capability” when this is the assessment.


Assessment Flow

1. Is the command channel encrypted or signed?
     No  → Tier 0
     Yes → Continue

2. Is GPS authenticated or navigation multi-modal?
     No  → Tier 1
     Yes → Continue

3. Anti-tamper, secure boot, LPI comms, adaptive decision logic?
     No  → Tier 2
     Yes → Tier 3

Real-world platforms may be mixed (Tier 1 comms, Tier 0 GPS). Phase D evaluates viability per attack vector, not per overall tier.


The Trajectory

The threat landscape is Tier 0 today. GPS authentication is deploying. Protocol signing is becoming standard. Multi-sensor navigation is commoditizing. The trajectory moves toward Tier 2 over the next 5-10 years.

ASTRA is designed for this. Fracture and Sever — the categories that survive hardening — exploit decision logic complexity and physical constraints that cannot be patched. Building expertise in these now creates capability that remains viable as the adversary hardens.


© 2026 Deep Woods Security. ASTRA is an open methodology. Use it. Extend it. Cite it.

This site uses Just the Docs, a documentation theme for Jekyll.