Join the crewMy track

Classes / Systems class / Systems Department

Systems Track: Spacecraft Links and Flight Software Basics · Module 1 of 4 · about 30 min

The link: how spacecraft talk to Earth

Every mission has two directions of traffic. The uplink carries commands from the ground to the spacecraft. The downlink carries telemetry back: health readings like temperatures, voltages and battery charge, plus the science or imaging data the mission exists to collect. When you hear an operator say they "have downlink", they mean the ground is hearing the spacecraft.

For missions far from Earth, NASA uses the Deep Space Network (DSN). It has three antenna complexes: Goldstone near Barstow, California; one near Madrid, Spain; and one near Canberra, Australia. They sit roughly 120 degrees of longitude apart, so as Earth turns, a distant spacecraft rises over one site before it sets at another. The largest dishes are 70 meters (230 feet) across. The DSN is run by JPL for NASA's Space Communications and Navigation (SCaN) program.

Distance is the problem that shapes everything else. Radio travels at the speed of light, and the gap between Earth and Mars swings from about 54.6 million km to about 400 million km. NASA puts the one-way delay at about 3 minutes at closest and about 22 minutes at farthest. Nobody joysticks a Mars rover. Teams send a planned sequence of commands, and the vehicle carries enough autonomy to handle the minutes in between.

Mars surface missions rarely talk straight to Earth for the bulk of their data. A rover has limited power and a small antenna, and it loses line of sight to Earth for about 12 hours a day as Mars rotates. Instead it hands data to orbiters overhead at rates up to 2 megabits per second, and the orbiters, with bigger antennas and more solar power, relay it to Earth. NASA calls this the Mars Relay Network, and says every image seen from the Martian surface since 2004 has come home through it.

NASA also counts link modes. If the ground is only receiving, the link is one-way. If the ground is sending an uplink that the spacecraft receives while its downlink is coming in, the link is two-way. Those words show up in operations jobs, so it helps to know them before an interview.

We separate three events: a command leaves Earth, the spacecraft receives it, and evidence returns. Radio rate changes how many bits fit in a pass, not the speed at which the first bit crosses space. A larger antenna may improve reception but cannot remove light time. We put generation time and Earth-receive time in different columns so a recent-looking dashboard update is not mistaken for a current measurement on Mars.

Worked example, fictional timeline: one-way light time is eight minutes, onboard handling is immediate, and other delays are ignored. A command sent at elapsed minute zero arrives at minute eight. Its immediate acknowledgement cannot arrive back before minute sixteen. That is a lower bound, not a promise: waiting for a pass, processing and queued traffic can add delay. Telemetry received at minute ten may describe conditions at minute two. We cannot use that measurement as proof of the spacecraft's exact state at minute ten.

NASA also distinguishes receiver lock from telemetry lock. A receiver can acquire the radio carrier while downstream equipment has not yet recognized the data frames. We check which layer reports success before declaring a usable stream. A received carrier does not prove that packets were decoded, that their timestamps are current or that a command finished. Each claim needs its own evidence, which is why ground monitor data and spacecraft telemetry are recorded separately.

For capacity, we declare units and assumptions. A fictional one-megabit-per-second useful data stream sustained for ten minutes carries 600 megabits, or 75 decimal megabytes. That calculation assumes the stated rate is useful data, excludes interruptions and does not establish the radio's physical capacity. If the rate instead includes overhead, useful payload is lower. We distinguish delivered volume from signal delay and from how long a station is available.

Practice, on paper: we draw a timeline for a different light time and add a five-minute scheduling wait. We label command receipt, earliest returned evidence and the age of a health reading. Then we make a pass-volume worksheet with units and an explicit overhead assumption. Our self-check requires both elapsed-time reasoning and a separate data-volume calculation. We keep this a fictional planning exercise, not a command sequence for a real spacecraft.

Sources

Module quiz

10 questions, 8 correct to pass. You'll see what you missed and why, and you can retake it.

Next: Decibels, antennas and the link budget →