Printable Version of Topic

Posted by: Greg Hullender Dec 20 2007, 07:04 PM

Given the recent news that Cassini will lose some data during the Phoenix landing due to contention for DNS access:

http://www.unmannedspaceflight.com/index.php?showtopic=4831&hl=

It seemed reasonable to create a topic to discuss where the DNS might be going.

Jasedm suggested a relay station at the Earth-Sun L2 point, which I think might be discussing (despite the expense) assuming the relay station could use laser communication with space probes, while beaming data to Earth via microwaves.

I know that experiments with laser communications were part of the now-abandoned Mars Telecommunications Orbiter, and I know that there was even a recent laser experiment involving Mercury Messenger.

http://www.space.com/missionlaunches/060104_laser_comm.html

These seem intended to support laser communications with ground-based receivers, though -- something that seems much too risky if remote space probes can't store days of data allowing for multiple retries.

Obviously a relay station would have its own risks -- you'd probably need two for redundancy -- and it'd be fabulously expensive, but it might also include quite a few savings as well. I've looked for figures for what the DSN costs to run, but I haven't found them yet. Not sure how large a part of a given mission (if any) is charged to the DSN, but it seems to me that a relay satellite ought to be a good bit cheaper to operate, and for far higher bandwidth.

Also, a laser transmitter ought to be lighter and consume less power than an equivalent microwave transmitter. That alone could result in huge savings for outer-system missions.

The only time I saw a serious proposal for a laser-to-satellite communications system was in a description of options for the "Grand Tour" that later became Voyager. (Assuming that counts as serious.) :-) With DSN needing more and more maintenance, I wonder if anyone is seriously considering laser for the future.

--Greg

Posted by: nprev Dec 20 2007, 07:29 PM

I proposed relay sats at the Earth-Sun L2 & L4 points here a couple of years back, Greg, but don't want to dig back that far. Turns out that somebody found that the idea dates from at least 1935(!) or thereabouts. Also, lasercomm from the outer solar system rapidly demands inordinate amounts of power due to inverse-exponential losses, so we're not quite there in terms of vehicle capabilities yet.

Not that I'm trying to shoot you down at all. I still really like the idea, think that even good old RF relay spacecraft at the Lagrange points would be immensely useful. The really nice part is that if you also set up a set of geosynchronous intermediate relay sats with crosslinks then you never need to care if the rain in Spain is falling mainly on the plain (to say nothing of the 70m dish)...

Posted by: Greg Hullender Dec 21 2007, 12:28 AM

Why would you have inverse exponential losses? One of the problems mentioned in the article I linked to was that the signal footprint is small enough to make aiming the laser challenging. Power dropoff should still be inverse-square, but the rate should be much, much lower.

As for locations, it occurs to me that geosynchronous orbit would probably be fine. The Earth subtends a fairly small area from there, so most of the sky is visible all the time, and interruptions would be short and quite predictable.

--Greg

Posted by: nprev Dec 21 2007, 01:29 AM

IIRC, most of the losses occurred as a result of beam-spread; there's considerable path divergence after several hundred million miles, so the receiver only gets a small fraction of the original transmitted energy. Could be wrong, though; that discussion was quite awhile back.

Posted by: tasp Dec 21 2007, 04:32 AM

Additional ground stations at, let's say, Samoa, Kourou, and Baikonour would 'fill in gaps'.

{I'm kind of dreaming of the day when we have so many space assets, this upgrade would be required}

Posted by: jasedm Jan 7 2008, 07:00 PM

Maybe a brief cloudburst over Madrid will be the least of our problems in the future. Having done a brief search, it appears that Earth-Sun L2 will get a tad busy in the coming years with the following planned missions to orbit there:

James Webb Space telescope
Terrestrial planet finder (may include multiple mirrors in close formation)
Darwin spacecraft (four or five separate free-flying components)
Gaia probe
Herschel space observatory
Planck surveyor

The WMA probe is already there observing cosmic background radiation, and probably many more missions wishing to take advantage of the small amounts of station-keeping propellant required, are in various stages of planning.
I'm not sure of the extent of the stable area available at L2, but it's going to get quite crowded (we may also find a few small but hazardous rocks there as well as the dust we know about)

Sooner or later I'm sure there will be some sort of solar-powered data relay at L2, especially as it could be maintained and upgraded by manned missions to mars which may use L2 rather than the moon, as a staging-point.

I think eventually the sheer number of demands on the DSN will make the space relay inevitable. Hopefully someone connected with the network reads UMSF and can enlighten us....

Posted by: elakdawalla Jan 7 2008, 07:14 PM

Space relays aren't the next step; large arrays of smaller dishes are. You might find my summary of http://www.planetary.org/blog/article/00000570/ to be informative. (The first half of that page is about the future of RTGs; skip halfway down the page to the paragraph starting "Which brings me to the next talk...")

--Emily

Posted by: djellison Jan 7 2008, 10:05 PM

That would include the replenishment of cryogenics to keep the receivers cool, the replacement of gyros (would gyros even work for a large dish?) and replenishment of thruster fuel to keep the think pointed.

It would cost an utter fortune, a fortune that would be much much better spent on array like DSN facilities on Earth, and upgrades to spacecraft that, with technology that is on the bench ready to go, take Mars-to-Earth comms up by two orders of magnitude over and above MRO.

Doug

Posted by: nprev Jan 8 2008, 03:17 AM

They would, but as sensors, not torquers like reaction wheels; you'd have to use an active RCS to keep everything aligned. It might work if we could fly manned servicing missions using the Constellation architecture, but, as you say, a serious cost/benefit study would be needed before making such a decision.

Posted by: Jim from NSF.com Jan 8 2008, 04:29 AM

Doesn't solve the problem, still need DSN or equivalent since the earth still rotates. So instead of sending assets into space, just add more dishes at the current DSN sites, they would be cheaper and last longer.

Posted by: Greg Hullender Jan 8 2008, 04:45 AM

Need maintenance be such a problem? Commerical communications satellites in geosynchronous orbit seem to sit there for years and years without it. And, as I mentioned earlier, Geosync is a much better place than the L2 point. (More sunlight there too.)

I do agree that putting radio dishes in space seems pointless, but I do still wonder if laser might not work much better. Dish size for sender and receiver is proportional to wavelength (if I recall correctly) so the size of what you have to orbit ought to be a great deal smaller than what you need on the ground.

I'm still thinking the power requirements for laser vs. microwave should be lower, not higher, since (again, if I recall correctly) it's a lot easier to keep a shorter-wavelength beam narrow. One of the articles on the topic even mentioned that targeting was a problem for laser, since the beam might still be just a few thousand miles across, as opposed to microwaves which are spread so widely that you only need to be roughly accurate.

Something that could reduce the weight and power requirements for all future probes would justify a good bit of expense, I'd think. Are my assumptions just wrong, or is there an additional factor that makes this impractical?

--Greg

Posted by: cndwrld Jan 8 2008, 07:44 AM

If money was no object, the best thing would be to add an additional 70 meter dish at each of the three current DSN sites. In my experience, we don't need new sites in different locations; we just have too much demand on the big dishes. ESA is building a full coverage system of 35 meter antennas, which is great. But the 70m dishes are just required for certain things. Getting ESA to pony up more cash for bigger dishes would be very helpful for the DSN, too. There is a lot of cross-support.

Posted by: djellison Jan 8 2008, 08:23 AM

Commercial comms sats and 70m DSN dishes are very very very different engineering challenges. I can see no way whereby money would be better spent on DSN facilities at L1/L2etc rather than on array facilities on Earth

Optical comms - currently there's no requirement for them. Perhaps once a vehicle uses it and a few ground stations start popping up, then maybe an on-orbit optical comms facility might make some sense - but - you still need ground facilities to which you would relay that data.

Doug

Posted by: edstrick Jan 8 2008, 09:41 AM

"...In my experience, we don't need new sites in different locations..."
Actually, we do.
We nees 6 DSN sites... North and south pairs. 120 degrees around the globe, roughly. There was an original S. African DSN site that was closed due to the anti-aparthide embargo, resulting in the building of the one in Spain. We've lost high-value encounter data at times due to the lack of a north-south redundant DSN pair during hardware failures or severe rain events, etc. Also, spacecraft at far south or north limits of the ecliptic <or beyond> can be harder to track from the opposite hemisphere.

Posted by: Doc Jan 8 2008, 10:42 AM

I believe that in the end the simplest thing to do would be to increase the number of DSN stations. They could build new ones at strtegic points in the southern hemisphere. There are hardly any there.

And by the way, has there been any development or consideration of installing a system which allows rapid switch of stations? For example; if one is down for some reason, another is linked to automatically to take its place. This would ensure steady uninterrupted downlink of data.

Of course this might need a little more stations to be available to be feasible.

Posted by: jasedm Jan 15 2008, 07:08 PM

Not trying to put you on the spot here Ed, but can you give examples? I'm trying to build up a picture of the percentage (cost) of a mission which could conceivably be written off due to deficiencies in the DSN.

Posted by: edstrick Jan 16 2008, 09:07 AM

My crusting over memory cells fail me at the moment. I recall some Cassini data lost due to outage and tracking conflicts. There was some significant data, but I don't recall what, lost from some mission a few years ago due to rain over Australia (I think) that an imaginary Japanese DSN station might have caught. But in general, a southern DSN station gets fewer high-quality tracking hours with a spacecraft high northerly ecliptic latitudes and vice versa, and rain events that muck up X-Band will RARELY occur at 2 different DSN stations at similar longitude but opposing latitude.

Mariner 10 did lose it's third encounter high rate data, planned around high resolution mosaics in the vicinity of "targets of interest" near closest approach, I believe due to a DSN breakdown. They had to switch to lower data rates and transmit only a vertical strip corresponding to the central 25% of each image which obliterated any contiguous coverage the spacecraft's wobbling attitude control would have permitted. The third encounter images are mostly forgettible, otherwise, it would have been as good as the other 2 encounters. The real science value was in the close in high latitude nightside flyby which confirmed the magnetic field was internal in origin, but that's partly due to the loss of 3/4 of the imaging data.

Putting a $$$ on it is worse than hard with arm-waving vague recollections, but we're stretched thin and the current 3 spacecraft in safe mode and one doing an encounter status point out just how thin is thin.

Posted by: nprev Jan 17 2008, 01:09 AM

Probably the only argument that really has a chance of working is financial. "We spend $X on these missions, and lose critical data due to DSN shortfalls and aging equipment, which grows more expensive to maintain each year. Therefore, we need to spend $X on upgrading the DSN in order to maximize the return on investment from these missions."

Gotta speak in beancounter in order to persuade beancounters...

Posted by: helvick Jan 17 2008, 09:23 AM

I've wondered before if the technologies and techniques that will be needed to build and manage the http://www.skatelescope.org/index.htm would make it easier (and cheaper) to build array based additional DSN capability. I'm also curious about how the SKA is being funded as it's not immediately clear from their web site where the €1.5bn that it is currently estimated it will cost to build will come from. Obviously there are many different engineering issues between building a science telescope and a deep space tx\rx ground station but it seems to me that there are many more similarities than differences and some sort of partnership between the SKA project and a global project to extend the DSN in terms of capability and redundancy would be very financially beneficial to both.

Posted by: Greg Hullender Jan 17 2008, 02:33 PM

I'm still thinking we just want to lobby for a "NASA Attachment" to ordinary commercial communications satellites. Something to sit on top, collect laser images from remote spacecraft, and send the data to NASA in the same Earth-bound datastream the satellite already manages.

Uh, I'm not proposing trying to add this to existing satellites. ;-)

--Greg

Posted by: djellison Jan 17 2008, 02:43 PM

A reception facility for optical comms isn't a 'bolt on' sort of resource - I don't have an exact figure, but we're talking metre+ scale light buckets.

Doug

Posted by: Greg Hullender Jan 17 2008, 05:48 PM

Didn't mean to trivialize the problem. Obviously there are a number of complications, including the fact that the optical part needs to be able to swivel to target different parts of the sky while the comm satellite always points "down". On the other hand, the comm satellite is very large -- maybe the side of a bus -- so we're probably not adding a large percentage to its mass. Also, the back part of the satellite (facing away from Earth) doesn't currently do much of anything, so an optical receiver on a boom extended away from the rest of the satellite shouldn't be in the way.

Best place would probably be on some successor to the TDRS L&K Satellites that NASA just hired Boeing to build. Given NASA is already paying to put up its own comm satellites in Geosync orbit, that would seem to be the logical place to put something like this. And the lifetime of the new satellites is probably 15 to 20 years, which is probably the right timeframe to evaluate optical comm systems.

--Greg

Posted by: dvandorn Jan 18 2008, 05:56 AM

I still have this image in my head of an antenna farm consisting of 10,000 old DirecTV and/or Dish Network 50cm dishes, all attached to altazimuth drives and being pointed by a central computer complex. Just lay the thing out on any convenient piece of desert or tundra, and run the multitude of signals received through rectifiers to pull out the weak signals coming from far away.

I'm sure there are good reasons why this would never work, but it seems like a good, cheap way to quickly put together a temporary addition to the DSN.

-the other Doug

Posted by: mcaplinger Jan 18 2008, 06:34 AM

Antenna arrays for the DSN have been considered for quite a while and I believe are now in active development.
For example, http://www.jb.man.ac.uk/ska/workshop/Weinreb.pdf

This certainly makes more sense that naive ideas of orbital facilities.

Posted by: edstrick Jan 18 2008, 08:20 AM

"...still have this image in my head of an antenna farm consisting of 10,000 old DirecTV and/or Dish Network 50cm dishes..."

The idea is not bad, but there's an optimum antenna size that balances antenna size, tracking mount size, and microwave components <feed horn, amp, whatever>. As pointed out above, antenna farm arrays are viable, but more in the scale of C-Band satellite dish, or small commercial dishes for communicating with synchronous sats.

Posted by: helvick Jan 18 2008, 09:05 AM

The expense that would result from the multiplication of the tracking, control, cooling and low noise tx\rx circuitry (for a start) would make an array of domestic satellite TV dishes very expensive very quickly. To get into the range of the signal to noise levels from a single 35m antenna you would need something like 4000 of the small home satellite TV dishes.

The SKA seems to be opting for 12-15m component dishes as that appears to be the economic sweet spot - you still need 34 of those to give you the equivalent of a single 70m dish (in terms of S/N ratio).

Posted by: edstrick Jan 18 2008, 09:11 AM

Exactly. And 34 of those -- with sufficient smarts -- can re-deploy to give lower aperture equivalents to 2 or 3 vehicles at once and provide more capability to handle more than one task at once.