Computer: Storage: Memory
Empty space as a storage medium   (+33, -2)  [vote for, against]
Juggle RF data packets for free data storage

If you can receive and re-transmit a data packet over the airwaves quickly enough, you can store data by tossing data packets at a reflective object (say, the Moon). The packet will take a couple of seconds to make the round trip, in which time many more packets could have been sent. The reflected packets are re-transmitted as soon as they're received.

You can keep many more packets in the air than you need to store locally, in much the same way that a juggler can keep more balls (bowling pins, chainsaws) in the air than he has hands. Each packet can be accessed once on each round trip. Reads are fast. Access time is slow, but hey, it's free.
-- HaltAndCatchFire, May 07 2006

Satellite Internet providers http://www.satsig.net/
This isn't quite Earth-to-Moon - but gives us ideas on bandwidth. [Jinbish, May 08 2006, last modified May 09 2006]

I’m sorry, but we lost all your data in the clouds.
-- ldischler, May 07 2006


Make sure you stick a post-it note on the plug asking the night-watchman not to use the socket for his portable tv.
-- 7ennyn, May 07 2006


Data storage devices (large mirrors sent floating off into orbit) would increase their storage capacity with time - as they drifted further away - thus automatically keeping up with current technology.

Massive mirrored extra-Kuipal croissant (so shaped to help focus and reflect incoming data) for you.[+]
-- zen_tom, May 07 2006


Only problem is that radio bandwidth is finite and generally already used for applications... Assuming we can find some available bandwidth for uplink and downlink (symmetrical bands, obvioulsy), we then have to consider the time the signal will take to travel to the moon and back (2 x 400000km approx./3 x 10^8 = 8/3 seconds) then the storage is equal to the effective bit-rate (Mbps) x 8/3...

Halfbaked to the max. (+)
-- Jinbish, May 07 2006


------------------\
+<---------------/
-- 2 fries shy of a happy meal, May 07 2006


Some questions:
How much storage can you fit into a meter? i.e. What is our expected bpm?(bytes per meter)

I suppose it depends on the frequency of the transmission medium - I'm imagining laser beams. We might need to consider using lower frequencies in order to combat dust and other particles that might get in the way (guessing that low-frequency EM radiation is less perturbed by dust, gas, clouds, bits of rubble etc)

One problem with this idea is that if something occludes the beam for any length of time, your data is lost.

One way to help reduce this problem might be to launch these mirrors perpendicular to the solar plane, so as to avoid any nasty planets or suns getting in the way and wiping out all our memory.

Likewise, if the system is switched off, you lose the memory again. And also, since the storage is maintained by juggling, and to keep data in the air, it needs to be recieved and then retransmitted, I can imagine that without some serious redundancy in the signal coupled with some equally serious checksum and auto-correction routines, we might very easily amplify any noise we get, and end up with the largest and most cosmic amp squeal ever created by mankind. Hendrix would be proud.
-- zen_tom, May 08 2006


//juggler can keep more ... chainsaws in the air than he has hands.//
... Or audience members/assistants in the vicinity.
-- Letsbuildafort, May 08 2006


Now there's an idea. With an array of mirrors so tilted as to zig-zag a beam along a well defined path (or have two parrallell mirrors, and introduce the beam an an angle only slightly off from perpendicular) you could (assuming minimal loss of signal per bounce) potentially create such a device in a small amount of space (which is kind of counter to the simplicity of the original idea)

Alternately, a ring of mirrors in space that send a signal orbiting around a central source (Earth), before beaming it back would give you a distance 2 x pi times the radius of the circle (plus 2 x that radius as the beam is sent out, and returned), increasing your storage by a fair amount indeed. (Wants to draw a picture)
-- zen_tom, May 08 2006


Byte per metre, by my back-of-envelope, would be the overall bit-rate used over the bandwidth - say 15Mbps for a DVB signal times 8/3 then divided by 8*10^8.... so that's about 5*10^-2 or half a bit per kilometre, or 1 byte every 16km.
-- Jinbish, May 08 2006


[zen tom] Periscopes as a storage medium?
-- fridge duck, May 08 2006


[miasere] - I don't think so -
//You can keep many more packets in the air than you need to store locally//
That's the untruth right there, at least with radio waves... but what if we used carrier pigeons! We could small, handwritten notes or maybe microfilm to the legs of an army of carrier pigeons and send them to their desitination. Once at the destination a new team could carry the information back.

That's not in the spirit of the idea, but it would have a greater storage capacity at approximately the same as, er... a load of microfilm.
-- Jinbish, May 08 2006


[fridge] Yes, lots of periscopes connected end to end - but I prefer the long distance, space version (less materials, and much cheaper once you've climbed up the gravity well)

If we're allowed 1 byte per 16km, that's 62.5 bytes per 1,000km or 62.5bpMm (Megametre) or 22.88KbpGm (Kilobytes per Gimametre) or, more cleanly 3.424Gb/AU (where 1 AU = 149,597,870.691 kilometers)

Sending out a ring of orbitals out to parking orbits on the inner edge of the asteroid belt at approx 2.7AU gives us a storage array of:

Outbound + Circuit + Inbound

Which equals 2.7 + (2 x pi x 2.7) + 2.7
= 42.5668AU

Which, at 3.424Gb/AU, gives us an overall storage capacity of 145.7486Gb - and a signal send/recieve time of around 5 and a half hours - which, to be fair, isn't that impressive.

How can we increase the amount of information coded into the beam? Is there a better alternative to DVB? And how is the signal transmitted? i.e. is there a specific type of modulation method that will serve our purposes best?
-- zen_tom, May 08 2006


DVB - Digital Video Broadcast, uses MPEG 2 transport and associated coding (OFDM modulation scheme - highest bit rate we got, I think, but requires tight synchronisation). Bit rates can be as high as 32Mbps, but the signal isn't as robust. It's used for terrestrial and satellite TV. It only made sense to suggest when talking about the Earth-to-Moon idea. We can use some other modulation scheme & EM bandwidth. I just don't have any numebrs, or idea, to hand. Besides, my link shows that renatable Internet style bandwidth via satellite ain't cheap!

We'll have to go further into halfbaked-ness to come up with a transmission medium/scheme... If we are bouncing off stuff from Earth then we need to use at least one device in geostationary orbits.
-- Jinbish, May 08 2006


Well, you could use masses of forward error correction and immediately delete what you've just sent; or perhaps use fountain codes... but you'd no doubt get the message back before you completely sent the original message... So it still doesn't make any more sense - but that's why I like it!
-- Jinbish, May 08 2006


cute in description, but you are basically describing any large networked system. The routers and firewalls on the net already perform this juggling act
-- theircompetitor, May 08 2006


[phlish] I'm not sure we need to store the data inbetween sending it. If we thought we might loose 3 packets in every 10 - we should use a (holographic?) coding algorythm that includes redundant data about other packets in each individual packet. If you fail to recieve some packages, they can be rebuilt using the data spread over the others. I believe this is what [JinBish] is referring to when he mentions forward error checking and maybe fountain codes. It means you have to accept some bloating of your data, the amount of bloat you accept is proportional to the robustness of your data signal.
-- zen_tom, May 09 2006


If you positioned your transmitter at a point appropriately close to a black hole, you'd be able to send off your data and have it whizz around, orbiting the gravitational anomaly in perpetuity, until you decided to scoop in and read it again. You'd need to fence off the area around the black hole so as to stop any unwanted light coming in and interfering with your signal. Different tangental angles of introduction would allow you to record information on different 'tracks'.
-- zen_tom, May 09 2006


[zen]: You've got it. Fountain codes basically represent the sent message spread over a whole load of transmitted packets. There isn't a need for re-transmissions or acknowledgement packets. You just keep sending the packets out and sooner or later you'll get a packet with the info you missed.... like someone standing with a cup next to a fountain. If you miss one drop you can just wait for the next 5.

However, the problem with fountain codes is that the sender needs to estimate when to stop sending out packets! So the point of reliability that [phlish] raises is still valid. The idea is flawed and won't work - but I like it's halfbakedness so much I'll try and argue to the contrary.
-- Jinbish, May 09 2006


Would it be possible to do something like this with a superconducting data cable formed into a loop?
-- Cuit_au_Four, May 11 2006


I've got jotted notes for the same idea from a few years ago. I never got a chance to post it here :-/

Anyway, to compat the obvious dangers (and low storage capacity), effectively a RAID would be needed. Multiple reflectors, some duplicating data and others adding storage. Relay around a series of satalites in synchronous orbit to increase the capacity.

Problem is, it just isn't worth it. How much does it cost to launch the reflectors into place, and how much does it cost to continually clean and re-transmit the reflected signals, compared to the ammount of practicle storage space? Take into account the inherent instability, and you got a thoroughly half-baked idea.

+
-- ironfroggy, May 12 2006


I like this idea. [zen_tom]'s spin reminds me of an older idea posted here in which by looking very close to a black hole, one can look at Earth because the outbound light from earth wraps clear around the hole.

There must be something smooth and reflective in space. Maybe some ice planet? It is just a matter of finding it. The farther away the better.

If nothing reflective can be found, another option would be to use high velocity charged particles as the medium. I am not confiednt they reflect very well, so they would have to impact something which then emitted em that we could detect.
-- bungston, May 13 2006


A better idea would be having the light orbit a tiny black hole in a slightly unstable orbit that would decay on certain intervals, and immediatley be reflected back into its orbit after being read. *smiles
-- DesertFox, May 13 2006



random, halfbakery