AutoTSN
A systematic approach from the first ideas to a perfectly configured TSN network.
Drag switches, ECUs, cameras and computers onto the grid and connect their ports. Every port takes exactly one cable; line rates from 100 Mbit/s to 10 Gbit/s.
Four steps to a schedule
- 01
Topology
Drag switches, ECUs, cameras and computers onto the grid and connect their ports. Every port takes exactly one cable; line rates from 100 Mbit/s to 10 Gbit/s.
- 02
Streams
Set talker, listener, frame size, period, maximum latency and jitter. AutoTSN lists every possible route and the best achievable latency before you schedule.
- 03
Schedule
Compute the schedule on the AutoTSN server with an exact SMT solver (Z3), or instantly and offline in the browser. A Gantt chart shows every link, and a table checks latency and jitter for every stream.
- 04
Gates and export
Every egress port appears with its eight gate channels. Move windows, fill gaps with best-effort windows and guard bands, and export as Linux taprio, NETCONF/YANG or JSON.
Three ways to compute
AutoTSN server
Exact SMT solver based on Z3. Finds a schedule whenever one exists, with up to 30 seconds per request.
Local engine
Zero-jitter heuristic scheduler in your browser, offline. Keeps frames of different streams apart in the egress queues and checks every result independently, but may miss tight solutions.
Relaxed with worst-case bounds
Leaves no stream out. Whatever fits gets exact windows, as with the local engine. The rest share a traffic class of their own in the gaps between them, and network calculus gives each of those streams an upper bound for latency and jitter. Where a bound is over the limit, you decide how to proceed.
Export formats
- Linux taprio
- tc commands for the time-aware priority shaper, optionally with hardware offload.
- NETCONF / YANG
- Gate parameter tables (ieee802-dot1q-sched) for NETCONF-capable bridges.
- JSON
- Every entry with start, duration and gate mask, for your own tooling.
For developers
The AutoTSN network model
How AutoTSN describes topology and streams, what it sends to a scheduler and what comes back. With an interactive example, a field reference and JSON Schemas, so you can use your networks with your own scheduler or plug your scheduler into the studio.
Your projects stay saved in your browser. Only when you choose the AutoTSN server are topology and streams sent to api.autotsn.de for the computation, or to the scheduler you entered yourself in Settings.