ECSS Communications — Short Contact Periods (space-systems/ecss/e50-short-contact-periods)
Use when the task is the short-contact provision of ECSS-E-ST-50C Rev.2 clause 5.6.3 — showing that the space link can move the mission data inside the contact periods the orbit and the ground network actually provide, including the short passes where link acquisition and setup take a visible share of the window.
Domain quick reference
- A contact period is not transfer time. Acquisition, carrier and symbol lock, and protocol setup are paid before the first user bit moves, and that cost is fixed per pass rather than per bit. The shorter the pass, the larger the share it takes.
- Two failure shapes follow from that. A pass shorter than its own setup delivers nothing and should be recognised as a non-contact rather than carried in the budget as a small one. A pass where setup takes most of the window still delivers, but the effective rate over the pass is far below the channel rate and sizing on the channel rate alone overstates it.
- The rate on the air is not the rate of user data. The coding rate and the framing efficiency each take their share, so the deliverable volume of a pass is the usable time times the channel rate times both factors.
- The budget closes over a plan, not over one pass. Between contacts the onboard store fills at the generation rate; during a contact it fills at the same rate while it drains at the delivered rate. The question is whether the backlog at the end of the plan is no larger than the backlog at the start.
- Storage capacity turns a backlog into a loss. Once the store is full, what is generated is dropped, and the dropped volume has to be counted rather than absorbed silently into a backlog number that looks stable only because the store stopped growing.
Workflow
- Validate each contact record: identifier, start time, duration, setup time, channel rate, coding rate and framing efficiency. A non-positive duration or rate, or a coding or framing factor outside its range, is an input error, not a case to clamp.
- Subtract the setup time from each pass to get the usable transfer time, flooring at zero rather than letting a negative time flow into the volume arithmetic.
- Convert the channel rate to a user-data rate through the coding rate and the framing efficiency, then multiply by the usable time for the deliverable volume of that pass.
- Raise the per-pass findings: a contact fully consumed by its own setup, and a contact where setup takes more than half the window.
- Order the schedule in time, refusing a duplicate identifier and refusing overlapping passes, which are a schedule error rather than a pass to merge.
- Walk the plan: accumulate generation between and during contacts, drain what the pass can deliver, clamp at the storage capacity and count what is lost when the store is full.
- Report the final and peak backlog, the lost volume, whether the store was ever cleared, and whether the plan is stable — that is, whether the link goes on carrying the mission across contacts this short and this irregularly spaced — absorbing representation error at the capacity boundary with a named tolerance.
Obligations
| Item | Step |
|---|---|
| ECSS-E-ST-50C Rev.2 5.6.3a | 7 |
Pitfalls
- Sizing a short pass on the channel rate. The effective rate across the pass is the deliverable volume divided by the whole contact period, and on a short pass the two differ by more than the link margin ever will.
- Treating a pass shorter than its setup as a small contact. It is a zero contact, and averaging it into a per-pass volume hides both the loss and the reason for it.
- Closing the budget on a per-pass basis. A pass can look sufficient while the plan still diverges, because generation continues between contacts and the gap is where the backlog is built.
- Reporting a stable backlog from a full store. A store at capacity stops growing because data is being dropped; the stable-looking number is the capacity, and the loss is the finding.
- Ignoring generation during the contact itself. On a long pass with a high generation rate, the data produced while the pass runs is a real part of what the pass has to carry.
Behavior contract (gate 3)
The contact validation, usable-time and effective-throughput computation, deliverable-volume and required-rate inversion, per-pass findings, time-ordering and overlap refusal, and the backlog walk with storage clamping are exercised by the gate 3 contract test: scripts/test_e50_short_contact_periods.py against scripts/e50_short_contact_periods_logic.py (stdlib unittest, offline). Run: python3 scripts/test_e50_short_contact_periods.py
Compliance
- ECSS standards are freely downloadable (ESA); cite the source and paraphrase per standards-map.yaml.
- compliance: STANDARDS-REF, gated: false.
