Data Volume Is Not a Data Rate
- A volume becomes a rate only after we know how much communication time is available
- Keep the units visible — they are often the first line of defense against nonsense
Lunar Orbiter Example
First Worked Example: A Lunar Orbiter
Given:
- 8 GB of science data each day
- Two scheduled 10-minute Earth contacts per day
- Assume every scheduled second can carry science data
What is the minimum data rate?
- Convert the daily volume:
- Contact time:
- Divide by the available contact time:
Only a Lower Bound
- 53.3 Mbit/s is only a lower bound; the real design must also cover:
- Protocol and coding overhead
- Acquisition and handover time
- Changing geometry and link performance (see a pass is not one range)
- Lost or shortened contacts
- Telemetry, commands, and competing users
- Variation in data production
Adding a Latency Need
- The orbiter also produces an event alert:
- Very small data volume
- Ten-minute delivery deadline
- Does the twice-daily contact schedule satisfy the mission?
- Total volume is trivial, but an alert generated between contacts sits on board until the next contact, hours later — far past the 10-minute deadline (lec01 slide 20)
- This is why the smallest data product may drive the most expensive design choice
Capacity and Availability Are Different
- Capacity asks: how much?
- Availability asks: when?
- Enough total capacity does not guarantee a usable path at the critical moment
- Timeline of scheduled service across a day:
- Scheduled contact → alert generated just after it → “no contact available — alert waits” → next contact
- Availability gets a precise, time-interval definition later: coverage becomes availability over time