NonaShield Intelligence
Device Identity Analysis

ONE SIGNAL
ISN'T IDENTITY.

A device ID, a User-Agent string, an IP address — each is spoofable alone. NonaShield fuses ~22 independent signals into a fingerprint anchored in hardware, not software claims.

// Baseline Device Fields

The SDK's device fingerprint collector reads 13 baseline fields on every session — zero dangerous permissions required.

Hardware ID
Android device ID
Build fingerprint
OS version / API level
Security patch level
Screen resolution
CPU core count
Total RAM
App version code / name
Package name
APK certificate fingerprint
Debug-build flag
Installer source

// Why Fields Alone Fail

  • Every field above lives in software — readable and forgeable by a rooted device.
  • Emulator farms fake all 13 fields in bulk, cheaply.
  • A fingerprint is only as strong as its weakest anchor.
The Hardware Anchor

Beyond the Software Fields.

Hardware-Attested Device Key

The SDK generates an ECDSA P-256 key in the hardware keystore with StrongBox-first, TEE-fallback logic, and reports which hardware tier actually backed it.

STRONGBOX TEE SOFTWARE (fallback)

Stable Device ID

Not a random UUID — the stable device ID is the SHA-256 hash of the device's own hardware-backed public key, formatted as a UUID. It changes only if the key material itself changes.

Stable ID = SHA-256(device public key)
// Rooted in silicon, not app storage

Device Anchor

Persistent SSAID (Android) / identifierForVendor (iOS) — flags factory-reset and device-swap attempts.

Wi-Fi Network Fingerprint

SHA-256-hashed SSID + BSSID — never raw network names — for "same network as before" comparison.

Emulator / Farm Fingerprint

Kernel device nodes, QEMU markers, build-property strings, and sensor-count anomalies fused into a HARD/SOFT evidence model.

The Verdict.

"Software fingerprints identify a claim. A hardware-attested fingerprint identifies a chip."

Hardware-Rooted Identity
Request Live Simulation →