I'm trying to learn about co-located GEO satellites. When I first heard about them, coming from a planetary orbital mechanics background, the first thing I pictured was based on the epicyclic approximation: the satellites are located on an epicycle that is centered on a GEO orbit. However, as I've read up a bit more, it seems that the satellites are all in GEO orbits (no extra eccentricity as with my epicycle picture), just phased very slightly differently.

Which of these (or neither) is the way that GEO satellites are co-located?

  • $\begingroup$ Actually I think I'll delete my answer and let someone else have the honors. GEO Satellites are station-kept by thrusters. See Wikipedia's Orbital Stability of GEO and Station Keeping. See also tagged questions and answers for geostationary and station-keeping $\endgroup$
    – uhoh
    Commented Sep 26, 2018 at 4:24
  • 1
    $\begingroup$ Currently I'm not sure what "satellites are located on an epicycle that is centered on a GEO orbit." means. Can you cite an example of a spacecraft that's been placed "on an epicycle"? So far all I've found is that the epicyclic approximation is a way to cope with axisymmetric terms without "doing the math" (numerical integration), useful when the potential is substantially axisymmetric or otherwise weird, and computing power is limited. What exactly would be an "epicycle" per se in this context, and how does one place an object on one of them? $\endgroup$
    – uhoh
    Commented Sep 26, 2018 at 4:36
  • 1
    $\begingroup$ It sounds like the OP was visualizing that the cg of the co-located sat system (centroid of the line connecting them for 2 identical sats) was centered on the line of the orbit with the system rotating around the cg. $\endgroup$ Commented Sep 26, 2018 at 10:20
  • 1
    $\begingroup$ @uhoh, what Organic Marble describes is indeed what I'm thinking. I can't find an online example of the epicyclic approximation being applied solely in a Keplerian case (which I realize is only an approximation of actual satellite orbits). The idea is, for small eccentricities, Keplerian orbits can be approximated as an object being on an epicycle (definition: "a small circle whose center moves around the circumference of a larger one") centered on the circular GEO. $\endgroup$ Commented Sep 26, 2018 at 21:44

1 Answer 1



In the old times, geostationary satellites had their longitude windows, each with around 1°, and tried to keep themselves at the center of this window. This strategy was good enough for its time, when few satellites with poor attitude controls were located in this orbit.

Nowadays, geostationary spacecraft are usually in a "collocation window", which is based on the windows of the past, but now more than one satellite is located at the same window. The basic technique to ensure two satellites are far from colliding is the eccentricity-inclination vector separation.

Basically, this technique gives a mathematical condition to impose for a modified set of elements, where each participant operator in the collocation window keeps the eccentricity vector and inclination vector in a circle around a nominal value.

Operators still need coordination with regards to RF (radio frequency) interference and procedures for collision avoidance maneuvers against dead satellites or debris. Luckily, these objects stabilise at a few degrees of inclination, hence they only pose a threat twice a year.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.