Kshana 0.35 tools
Trust, evidence and test tools
The 0.35 features that run in your browser, and where to find the ones that cannot. Files you open here are read in this page and never uploaded; the engine runs on your computer as WebAssembly.
Advisory software. Kshana is not type-approved navigation equipment (IEC 61108, IEC 61162) and has not been assessed against any performance standard. Its scores, authentication status and exports are aids to judgement and records of what the engine computed. They are not a certification, a legal opinion or a finding of fact about any event.
Also in the Studio: Geometry, the skyplot, skyline and dilution of precision for a place and a time; Lunar station, the modelled coverage of a lunar reference station; Maritime trust; Interference map; Training debrief.
The engine starts when you use a tool.
Every 0.35 feature, and where it is
In this browser where it can run. The rest are command-line or server features, and the reason is stated.
| Feature | In the Studio | Also |
|---|
Verify an evidence pack
An evidence pack is a signed technical record of what the engine computed from a receiver log. It is not a legal opinion. Choose the files of a pack (manifest.json, manifest.sig, log-slice.bin, config.json, epochs.json, summary.html, and timestamp.tsr if there is one). Packs are made with the command line, Python or the AI assistant tools: signing needs a private key, which should not pass through a web page.
No pack files chosen.
Attach a timestamp token
An RFC 3161 token binds a time to manifest.json. Request it with openssl ts -query, send it to your timestamp authority, then attach the reply here. This does not check the authority's signature or certificate chain: run openssl ts -verify for that.
Export a scenario for a laboratory simulator
Kshana writes the vehicle motion, the NMEA, the waypoints and the event times. It writes no signal: nothing here synthesises or transmits a radio-frequency or baseband waveform, and an event is a labelled time interval, not a recipe for interference. Run a simulator only where that is authorised.
Which public-framework rows do your runs support evidence for?
Kshana outputs support evidence for a requirement. They do not certify a product, show that it complies with a framework, or make it conformant. Kshana is not a conformity-assessment body. Requirement text is paraphrased, never reproduced. Choose result files (*.result.json) and, optionally, the scenario files beside them (*.toml, matched by name).
No result files chosen.
Galileo OSNMA: is the navigation data authenticated?
Load a recorded u-blox UBX-RXM-SFRBX file or a plain I/NAV page file and see, per satellite and per tag, whether the data is cryptographically authenticated and why a check failed. Or open the JSON that kshana osnma verify --json wrote. Synthetic and official-vector samples only.
Advisory. This is analysis and monitoring output and not type-approved navigation equipment. The authentication status reported here is not a navigation integrity service and must not be the sole basis for any navigation or safety decision. Time: checks are relative to the time stamps in the pages. Authentication says the data came from the system; it says nothing about the signal it arrived on or the quality of the fix.
Nothing open yet.
IMU fit: sensor parameters for the hybrid and fusion scenarios
The hybrid and fusion scenarios are simulations that take sensor noise parameters, not a measured series. This reads those parameters off a stationary IMU log (the overlapping Allan deviation of the integrated signal) and writes a ready-to-paste [accel_classical] table that names the log's SHA-256. The file is read in this browser and never uploaded.
MODELLED, advisory. The estimator is checked on synthetic white-noise logs with known coefficients (recovered within 10 %), not against an external oracle. A log that is not still, under 2000 samples or unevenly sampled is refused or gives parameters the Allan slopes flag.
Nothing open yet.
Jammertest: score a receiver log against the published test plan
The Norwegian Jammertest publishes a test catalogue and a transmission plan. Kshana reads them as scheduled truth, the slots the plan says, not recorded on-air times, and scores a receiver-trust run against them: per test, was it detected, how late, did it recover, and which alarms fell outside every slot. Choose the catalogue and plan files you downloaded yourself (Kshana does not bundle them) and a receiver log. Everything runs in this browser; nothing is uploaded.
MODELLED, advisory. The slots are the plan's, so a detection probability is a lower bound; no external oracle states when a transmitter was on. A scorecard describes one log, never a product. Signed score packs are made with the command line (they need a signing key).
Nothing open yet.