Local Signals#
Network‑Based Signals#
Network behavior is one of the strongest permissionless indicators of environmental conditions. The RTT‑App observes patterns that emerge during routine requests.
- DNS Resolution Variance — fluctuations in lookup time indicate instability or congestion.
- TLS Handshake Duration — longer or inconsistent handshakes suggest degraded paths or interference.
- QUIC Handshake Success and Fallback — repeated fallback to TCP/HTTP/2 reflects drift or loss.
- Protocol Negotiation — unexpected downgrades signal environmental inconsistency.
- Retry Patterns — increased retries imply packet loss or transient instability.
- Captive Portal Detection — redirects or blocked requests indicate constrained environments.
These signals require no special permissions and behave consistently across platforms.
Device‑Stability Signals#
The app can also observe local stability through standard lifecycle and runtime indicators.
- UI Thread Jank Metrics — frame drops and animation stalls reflect local load or contention.
- Memory Pressure Warnings — low‑memory conditions indicate local drift or resource strain.
- Thermal Throttling (Android) — thermal events reduce performance and signal environmental stress.
- Event Loop Jitter — irregular scheduling suggests instability in the local runtime.
- Foreground/Background Transitions — rapid or unexpected transitions can correlate with drift.
These signals do not reveal personal information and do not require system access.
Signal Characteristics#
Local signals share several important properties:
- Permissionless — all signals are available without user prompts or system privileges.
- Coarse but Reliable — they reflect environmental conditions without fine‑grained detail.
- Platform‑Stable — behavior is consistent across iOS and Android.
- Non‑Diagnostic — they indicate drift or clarity but do not identify root causes.
- Composable — each signal contributes independently to the Awareness model.
These characteristics make local signals ideal for RTT‑App v1.
Role in the Awareness Model#
Local signals form one half of the Awareness computation. They describe how the device “feels” right now. When merged with the server‑declared state, they produce a unified Awareness state that reflects both local and global conditions.
