ComNet Ideas for maximum coverage and minimum upkeep

[Leafbaron]

Messing around with the comnets yesterday I realized that setting up a LKO or geostationary comnet is not necessary. The ground stations on Kerbin are extremely powerful and the only way you would lose connection with Kerbin is if you were on the far side of the Mun or Minmus. So my thought was to build a comnet with 60 degree phasing from Minmus and the Mun on both sides, I believe these are called Lagrange points but I'm not positive. I just feel that this would give you optimal coverage on the fariside of Kerbin's Moons with the least amount of infrastructure design. Please let me know your thoughts, I'm new to developing a ComNet.

 

Thank You,

 

-Leafy

[regex]

Yep, just like IRL you only need relays for specific situations, like making sure your Duna rover always has signal. There is no need to create a massive relay of satellites to ensure coverage. I've also noticed that launch times are so short, and Kerbin so small, that there is no need for a launch relay because your primary station is always in view (launch sats or ensuring you have stations along the launch route are vital in RO/RSS).

[blowfish]

10 minutes ago, Leafbaron said:

I believe these are called Lagrange points but I'm not positive

Lagrange points are something entirely different, which does not exist in stock KSP (because you need multi-body gravitation).

[Leafbaron]

6 minutes ago, blowfish said:

Lagrange points are something entirely different, which does not exist in stock KSP (because you need multi-body gravitation).

I see, so then it would just be 60 degree phasing since KSP doesn't model multiple gravitational forces like this

as far as full communication coverage would the 60 degree phasing work?

Edited September 15, 2016 by Leafbaron

[blowfish]

16 minutes ago, Leafbaron said:

as far as full communication coverage would the 60 degree phasing work?

120° phasing (3 satellites) should be sufficient provided they are high enough (although there would probably still be coverage gaps near the poles).  It's important to match up the orbital periods very closely though (more than any other orbital parameter) - small differences can add up over many orbits and lead to gaps in your coverage.

[Sharpy]

A single satellite trailing or leading the occluding body by some not too excessive value (60 degrees is okay) is all that's needed. So, yes, Kerbin-Mun L4 and L5 are okay points.

Alternatively, two satellites at an arbitrary not-too-low orbit, 90 degrees from each other (but far enough that. When Mun occludes both Kerbin and one of the sats, the other is still on the same side, not occluded by Kerbin. If both are occluded by Mun, Kerbin is in sight. Just make the orbit about 2x Kerbin radius, so that no situation when both Kerbin and one sat are occluded, and the other is already below Kerbin horizon.

Another two relatively inexpensive alternatives: two polar orbits, crossing at 90 degrees; arbitrary. If one is occluded along with Kerbin, the other is always in sight. And one very neat though somewhat harder option: highly inclined, eccentric orbit resonant with Mun.

The idea is your satellite spends most of the time "above" Kerbin's north pole or these regions. By making the orbital period a constant fraction of Mun's orbital period, you make sure it's in the same location of orbit whenever an eclipse happens. Adjusting the phase you may make sure that location is well above Kerbin, near apoapsis. That way your sat is always visible "above" Kerbin whenever its orbit could be occluded.

Edited September 15, 2016 by Sharpy

[Leafbaron]

29 minutes ago, blowfish said:

120° phasing (3 satellites) should be sufficient provided they are high enough (although there would probably still be coverage gaps near the poles).  It's important to match up the orbital periods very closely though (more than any other orbital parameter) - small differences can add up over many orbits and lead to gaps in your coverage.

60 degrees either side of the mun or minmus would be 120 degree phasing if the observer was focused on kerbin. So essentially 120 degree phasing with one of Kerbin's moon smack dab in the middle of the two orbiting satellites

[blowfish]

3 minutes ago, Leafbaron said:

60 degrees either side of the mun or minmus would be 120 degree phasing if the observer was focused on kerbin. So essentially 120 degree phasing with one of Kerbin's moon smack dab in the middle of the two orbiting satellites

I meant orbiting around mun/minimus.  If the satellites are orbiting Kerbin then you're still going to have a coverage gap.

[Sharpy]

2 minutes ago, blowfish said:

If the satellites are orbiting Kerbin then you're still going to have a coverage gap.

Absolutely not. If the satellite is trailing Mun by 60 degrees, there is no situation when Mun can simultaneously occlude the satellite and Kerbin.

[Leafbaron]

23 minutes ago, Sharpy said:

Absolutely not. If the satellite is trailing Mun by 60 degrees, there is no situation when Mun can simultaneously occlude the satellite and Kerbin.

And if orbiting the mun and you have a satellite leading at 60 degrees and trailing at 60 degrees you should have coverage orbiting on the far side of the mun. correct?

[blowfish]

10 minutes ago, Sharpy said:

Absolutely not. If the satellite is trailing Mun by 60 degrees, there is no situation when Mun can simultaneously occlude the satellite and Kerbin.

A single satellite trailing Mun would lead to a ~120° coverage gap up to about 200 km on the side of Mun farthest from the satellite and from Kerbin.  If you had another satellite leading Mun by 60 degrees, the gap would be about 60° and only extend up to about 30 km.

[DerekL1963]

34 minutes ago, Sharpy said:

The idea is your satellite spends most of the time "above" Kerbin's north pole or these regions. By making the orbital period a constant fraction of Mun's orbital period, you make sure it's in the same location of orbit whenever an eclipse happens. Adjusting the phase you may make sure that location is well above Kerbin, near apoapsis. That way your sat is always visible "above" Kerbin whenever its orbit could be occluded.

As someone with extensive experience with RT..  the problem within Kerbin's SOI isn't generally north-south, it's east-west.   That's why relay birds at the (simulated) L points are so popular, they allow the network to "see" behind the Mun.  (Which was Leafbaron's original question - "I just feel that this would give you optimal coverage on the fariside of Kerbin's Moons with the least amount of infrastructure design".)  North-south birds are useful for longer (Kerbol system) range relays in RT because (with an equatorial belt of shorter ranged relays) they avoid the problem of KSC being occluded.  However, with CommNet's equatorial ground stations you no longer have that problem, if you can see Kerbin and have an antenna with sufficient range, you always have a connection to KSC.

CommNet has some significant differences from RT, and because of that you need to rethink the logic behind your network design from the ground up.

[Sharpy]

12 minutes ago, blowfish said:

A single satellite trailing Mun would lead to a ~120° coverage gap up to about 200 km on the side of Mun farthest from the satellite and from Kerbin.  If you had another satellite leading Mun by 60 degrees, the gap would be about 60° and only extend up to about 30 km.

This can still be solved with three Kerbin-centric satellites - Molinya-style; in 1:1 resonance with Mun but orbit eccentric enough that the satellite remains "above" until the other takes over. Though true, orbiting Mun is simpler.

[eddiew]

Just my 2p, but any attempt at a precision orbit in KSP is a waste of time. Put it at Munar altitude, give it the same orbital period to the second... and it'll turn out to be 0.4 seconds more or less. Over a couple of decades (say a trip to Jool and Eeloo), the Mun will eat your satellite, and you'll only find out when you need it right now 

This is casting back to my RemoteTech adventure in 0.25, but the best luck I had was with a flower formation like this:

I suspect you could place your relays in a polar orbit, with an AP between Mun and Minmus, and they would never be captured  

Re covering the back of the Mun... put a couple of satellites in Munar orbit and combine them with the flower. You can quickly create a 'high probability' that you will have a satellites in the right place at any given time. Trying to guarantee that you always do will severely fray your nerves

 

[Leafbaron]

Thanks for the input @eddiew!

[eddiew]

No probs @Leafbaron, tis a diagram I'd put together a while ago, thought it might spare you some pain

Not that fixed (e.g. lagrange/geostationary) satellites aren't possible, but if you choose to do that, prepare to inspect them every year or two for station keeping duty... And you better have very delicate thrust on them, e.g. RCS+capslock ^^

[linuxgurugamer]

1 minute ago, eddiew said:

No probs @Leafbaron, tis a diagram I'd put together a while ago, thought it might spare you some pain

Not that fixed (e.g. lagrange/geostationary) satellites aren't possible, but if you choose to do that, prepare to inspect them every year or two for station keeping duty... And you better have very delicate thrust on them, e.g. RCS+capslock ^^

Use an ant engine, throttleable  and weak thrust

[Leafbaron]

5 minutes ago, eddiew said:

Not that fixed (e.g. lagrange/geostationary) satellites aren't possible, but if you choose to do that, prepare to inspect them every year or two for station keeping duty... And you better have very delicate thrust on them, e.g. RCS+capslock ^^

Exactly what I planned on doing!  no way to get "precision" orbits without RCS and CAPS lock that I've found. Hopefully B9 aerospace updates to 1.2 soon because I love the rcs blocks it has and if I remember correctly if you tap the translate keys with b9 you get very small thrust, almost like tapping the throttle button and if you hold it down it throttles up to full power. 

[DerekL1963]

22 minutes ago, eddiew said:

Just my 2p, but any attempt at a precision orbit in KSP is a waste of time. Put it at Munar altitude, give it the same orbital period to the second... and it'll turn out to be 0.4 seconds more or less. Over a couple of decades (say a trip to Jool and Eeloo), the Mun will eat your satellite, and you'll only find out when you need it right now 

Using MechJeb or Engineer Redux to monitor your orbital period, with a little practice (and tweaking your source of thrust to the absolute minimum) it's quite possible to match orbits within .01 second.   Once you've done that (for Munar Lagrange point relays), I suspect you'll only need to check it once a century or so.
 

24 minutes ago, eddiew said:

This is casting back to my RemoteTech adventure in 0.25, but the best luck I had was with a flower formation like this:

I said it above, and I suspect I'll be repeating it a lot in the coming weeks, CommNet is not RemoTech.  Your "flower" is designed to solve a problem that CommNet doesn't have - that of KSC being in 'shadow' because it's on the wrong side of Kerbin from your satellite's point of view.  CommNet provides an equatorial belt of ground stations, so basically if you can see Kerbin you have connectivity to KSC.  With CommNet and Munar Lagrange point relays, your "flower" is completely superfluous as CommNet will simply connect directly to one of those ground stations.

[eddiew]

@DerekL1963 - I did note above that for these purposes, tilting the orbits 90 degrees and setting their APs between Mun and Minmus would be sensible. For the OP's case, I think it would give 'a good chance' (TM) that one of them would be far enough out from Kerbin to talk to a satellite on the back side of a high Munar orbit - without needing too much precision. It would also give 'a very good chance' that Mun would never occlude Kerbin when connecting from an interplanetary target  

On the other hand, if you're confident to the microsecond, then fair enough   My comsat strats were developed in 0.25 when the game was new to me, and I hadn't really availed myself of all the tools available to help. But I do have a strong memory of how painful that learning curve was and felt it worthwhile pointing out that perfect orbits need a very meticulous player. Turns out I am not one ^^;

[Kerbart]

Two polar orbits on a 90° relative inclination. Give them a decent altitude–1000km will do–and one of them will always be visible from any angle (barring solar conjuctions and the mun interfering). I used this setup as my "second tier" in Remote Tech, with a low orbit first tier of four satellites (90° spaced in an equatorial orbit) to provide a connection to the ground station. Since connecting to the ground station is not a problem, this setup should do just fine.

[septemberWaves]

2 hours ago, eddiew said:

Not that fixed (e.g. lagrange/geostationary) satellites aren't possible, but if you choose to do that, prepare to inspect them every year or two for station keeping duty... And you better have very delicate thrust on them, e.g. RCS+capslock ^^

I get the feeling that when I'm setting up late-game communications networks, I'll be using ion propulsion as my ACS. It'll mean I have to set up very precise maneuver nodes, but that won't really be a problem. Alternatively, I'm pretty sure thrust-limiting RCS thrusters is doable in stock, so that's another option for the early game.

[RocketSquid]

Constant ground contact is actually unnecessary. You just need a relay with a relay antenna (RA-5, RA-50, RA-100, or HG-5) on a satellite that has near-constant contact with your probe and occasional contact with Kerbin (say, five minutes once a day). If you send to it while it has contact, then it will simply relay it. If it doesn't have contact, it will save your data for later. I'm not too sure, but the part descriptions on the RA series points to the antennae having the ability to store data for later transmission.

[LameLefty]

2 hours ago, RocketSquid said:

Constant ground contact is actually unnecessary. You just need a relay with a relay antenna (RA-5, RA-50, RA-100, or HG-5) ...

This is a big point that has so far gone unmentioned here - CommNet has some specific limits - you have to have a Relay antenna, not just any ordinary antenna, to relay control. And right now, Relay antennas are fairly high up the Tech Tree - no uncrewed far side Münar probe landings or precision maneuvers until you unlock them.

[RocketSquid]

33 minutes ago, LameLefty said:

This is a big point that has so far gone unmentioned here - CommNet has some specific limits - you have to have a Relay antenna, not just any ordinary antenna, to relay control. And right now, Relay antennas are fairly high up the Tech Tree - no uncrewed far side Münar probe landings or precision maneuvers until you unlock them.

You're forgetting the HG-5. It's unlocked at Basic Science, so the same time as the Stayputnik, and has rudimentary relay functions and a range reaching past Munar orbit.