Example: What a network looks like with no ground stations

[ansaman]

Satellites 1, 2, and 3 are in Keosynchronous orbits roughly forming an equilateral triangle. The backup is always in view of at least one satellite and most often two.  KSC is always in view of two satellites. This gives coverage out to the range limits of the system.

 

 

Edited September 20, 2016 by ansaman

[KerikBalm]

Yea... but this requires fairly powerful relay sats. The HG-5 won't cut it.

[Sereneti]

you dont need a keostatnionary orbit for that...

its enough if one satelite see the two other satelites (same triangle with lower orbit) and one of this satelites see the KSC...

Keostationari are good, if you have to aim with antennas, but thats not the case...

 

[_stilgar_]

2 hours ago, KerikBalm said:

The HG-5 won't cut it.

You can stack multiple HG-5s

[severedsolo]

41 minutes ago, _stilgar_ said:

You can stack multiple HG-5s

Not any more. THe last build turned off stacking for all but the Communotron 16.

[KerikBalm]

When did that build come out?! No!!!!!!

So much for my deep space relay for use with OPM

[ansaman]

3 hours ago, Sereneti said:

you dont need a keostatnionary orbit for that...

its enough if one satelite see the two other satelites (same triangle with lower orbit) and one of this satelites see the KSC...

Keostationari are good, if you have to aim with antennas, but thats not the case...

Good point, but with a lower orbit (and with settings for strict Line of Sight) there might be gaps in connectivity to KSC. It would be a good project or math problem to calculate just how low an orbit configuration would still allow continuous connection with KSC. Of course, a fourth satellite would help fix that. I am betting that for each additional satellite, the minimum orbit is lowered.

[DerekL1963]

4 hours ago, ansaman said:

Good point, but with a lower orbit (and with settings for strict Line of Sight) there might be gaps in connectivity to KSC.

There isn't.  I used to use an array of relay birds in a 1000km orbit in RemoteTech (which enforced strict LOS) all the time.  To eliminate gaps there's only three requirements; a) that all birds be in LOS to each other. b) that they have the range to reach each other,  and c) that at least one bird have LOS to KSC.
 

4 hours ago, ansaman said:

It would be a good project or math problem to calculate just how low an orbit configuration would still allow continuous connection with KSC.

For maintaining LOS to each other, the RemoTech visual planner suggests the lower bound is a little below 700km.  You'd have to test to see if an array that low will maintain LOS to KSC.  IIRC my low array in RemoTech required four birds, but I can't recall if that was to maintain connectivity or to increase the array's tolerance to orbit drift.

Kerbosynchronous orbits and arrays with three birds are popular because they let the array builder show of his l33t skills, and that's how it's done in the Real World - but there's no physical reason that limits you to such arrays.  I've built arrays with very high connectivity that had no birds in the same orbit at all in RemoTech.

The HG-55 will reach out to the Mun and beyond, it should more than suffice for a Kerbin orbital relay system.  In fact, under the current build, it's practically the only use for it.

[Sereneti]

its good to have one satelite on a Keo-stationary orbit over the KSC. 
that is enough to reach nearly the half -site from Kerbin. ( a white point is over the top and down of kerbin)

that make the whole thing a lot more easy.
it avoides to have gaps in the LOS between the satelites and KSC.

[Nicias]

If my napkin math is correct, if you have n satellites, the altitude so that they can always see each other is R[sectant(Pi/n)-1], (and hence all of the equator including KSC,) where R is Kerbin's radius. So (rounding up to the nearest km)

Satellites	Altitude
3	600km
4	249km
5	142km
6	92km

 

Seven would give you 66km, so inside the atmosphere.

[juanml82]

Wouldn't powerful satellites in polar orbits work best? I haven't played too much with the new comm system, but I guess a couple of powerful satellites in a highly inclined solar orbits should be able to avoid most instances of occlusion

[rcp27]

When I dabbled in remote tech in the past I favored 4 satellites in 1000 km circular equatorial orbits because that would provide effective coverage within the range of a communotron 16 in RT. Going for a synchronous orbit brought hassles involving directional antennas in RT (not such a problem with the new stock implementation).  Playing around with RT I personally found the easiest way to handle unmanned launches was to have something in low orbit and time my launch for when it passed overhead so that it could serve as a relay in case my gravity turn was too shallow to deploy a stronger than dipole antenna before the KSC dropped over the horizon. The current stock implementation is much more forgiving than RT, though (and rightly so, IMO).

[regex]

14 hours ago, ansaman said:

Is this sort of gamey Remote Tech thing actually required in the vanilla system or are you just removing ground stations to make things "hard" for yourself?

[DerekL1963]

1 hour ago, juanml82 said:

Wouldn't powerful satellites in polar orbits work best? I haven't played too much with the new comm system, but I guess a couple of powerful satellites in a highly inclined solar orbits should be able to avoid most instances of occlusion

Birds in polar orbits are only in LOS to KSC very rarely.  They're best used (in my experience) with 1-2 equatorial birds.

 

23 minutes ago, regex said:

Is this sort of gamey Remote Tech thing actually required in the vanilla system or are you just removing ground stations to make things "hard" for yourself?

The new [stock] CommNet system (which can be disabled entirely if you wish) has an option to disable the additional DSN stations, leaving KSC the only ground station.

[ansaman]

33 minutes ago, regex said:

Is this sort of gamey Remote Tech thing actually required in the vanilla system or are you just removing ground stations to make things "hard" for yourself?

I felt that the ground stations made it a little easy early in the game and I wanted to see what it would take. Obviously by some of the discussion there are different ways to skin the cat.

2 hours ago, Nicias said:

If my napkin math is correct, if you have n satellites, the altitude so that they can always see each other is R[sectant(Pi/n)-1], (and hence all of the equator including KSC,) where R is Kerbin's radius. So (rounding up to the nearest km)

Satellites	Altitude
3	600km
4	249km
5	142km
6	92km

 

Seven would give you 66km, so inside the atmosphere.

You get a gold star!    

[regex]

15 minutes ago, ansaman said:

I felt that the ground stations made it a little easy early in the game and I wanted to see what it would take.

Yeah, that's old Remote Tech behavior. We don't even do that in RO, although that doesn't mean you don't have to put up relays.

[Rayder]

My math got 600km for the lower limit too. Any lower and one will be over the horizon. It's worth noting that at this altitude, your triangle would have to be spot on. A small deviation will put one of them below the horizon of another.

On a related note, I've been working on a design for some time that provides 100% uptime (or very, very close to it...) and requires no orbital maintenance. My design is two satellites in high polar orbits perpendicular to each other. A third satellite sits in an equatorial orbit far enough away so that one satellite is always visible to the relay. The relay also serves to route signals between the polar satcoms in case one is on the blind side.

In this configuration, above the blind zone you will always have vision to one polar satcom. A polar satcom altitude of 9,000 km leaves a dead zone of about 300km. Unfortunately the values are inversely exponential, and to get a lower dead zone your DSN altitude goes up very high. The lowest dead zone I can achieve with this setup is 250km.

I'm yet to practically test this design out, but I believe it's got great uptime. Any downtime would likely be caused by the swiss cheese effect.

 

[Sharpy]

11 hours ago, DerekL1963 said:

Birds in polar orbits are only in LOS to KSC very rarely.  They're best used (in my experience) with 1-2 equatorial birds.

 

Polar, strongly eccentric.

One geostationary, to relay from KSC, one or two polar of very high eccentricity. With one, you're "blacked out" for maybe 10 minutes in a week. With two, only if some other body than Kerbin occludes both (or occludes one, while the other is on its 10 minutes in a week periapsis pass.)

Edit: Of course Kerbin coverage will be poor for the opposite hemisphere, unless you add another polar with apoapsis above the opposite pole.

Edited September 21, 2016 by Sharpy

[RizzoTheRat]

I always use 4 LKO sats in RemoteTech as rcp27 describes above.  Think it's about 780km to get an orbital period of 90 minutes.  This means I have permanent coverage on the surface all round the equator and later in the game can build my own ground stations for long distance comms that communicate with KSC via the LKO sats.  The advantage of this over Keosynchronous sats is they're much closer to the center of kerbin and therefore can be targeted with narrow angle dishes at a shorter range (not sure if this is an issue in the new stock implementation)

[DerekL1963]

4 hours ago, RizzoTheRat said:

The advantage of this over Keosynchronous sats is they're much closer to the center of kerbin and therefore can be targeted with narrow angle dishes at a shorter range (not sure if this is an issue in the new stock implementation)

No, it's not an issue.   CommNet (the new stock system) only checks to see if there is LOS.