Skip to content

arsubsv-field-mode-p4.md

Latest commit

 

History

History
47 lines (33 loc) · 2.07 KB

arsubsv-field-mode-p4.md

File metadata and controls

47 lines (33 loc) · 2.07 KB
Raw

ArSubsv Field Mode P4

Field Mode is the practical receiver workflow for live IEC 61850 Sampled Values and offline PCAP/PCAPNG analysis.

The P4 application branch is rebuilt directly on merged main and pins the field-core engine by its exact commit SHA. The draft PR therefore contains only Field Mode changes and remains safe to test locally without merging.

Five evidence axes

ArSubsv separates:

  • CAPTURE: frames available to the application;
  • PROTOCOL: Ethernet/SV APDU decoding;
  • STREAM: sample-counter continuity and payload consistency;
  • CONFIGURATION: observed traffic versus bound CID/SCD;
  • MEASUREMENT: channel semantics, scaling, CT/VT context, and signal confidence.

A configuration mismatch does not make a clean protocol stream BAD. Unknown measurement semantics remain UNKNOWN rather than being presented as amperes or volts.

Offline workflow

  1. Open .pcap or .pcapng.
  2. Select a discovered stream.
  3. Review raw decoded values and continuity without SCL.
  4. Import .cid, .scd, .icd, or .iid when available.
  5. Review scored SCL binding and expected-versus-observed findings.
  6. Enter explicit CT/VT context only from reviewed evidence.
  7. Export a support bundle for reproducible review.

Support bundle

The initial P4 bundle is metadata-only by default and contains:

  • manifest and SHA-256 checksums;
  • receiver evidence in Markdown and JSON;
  • five-axis field summary;
  • selected-stream diagnostics;
  • measurement context when configured;
  • application and engine revision evidence;
  • SCL hash when an SCL file is loaded.

The full capture and original SCL are not copied silently. Capture-excerpt and privacy-selection UI are planned as the next tranche.

Current limitations

  • Known-injection comparison exists in ARIEC61850 but the ArSubsv entry dialog is not yet implemented.
  • Quiet/noise-floor classification is evidence-based and non-destructive, but waveform minimum-axis presentation still needs local UI review.
  • Real Ahmed Mohaisn PCAPNG/CID replay remains a local acceptance gate; deterministic tests do not replace real-device evidence.