Boffins' satcomms rig uses earthly LEDs to talk to orbiting PV panels As low-cost satellites become more common, researchers are turning their attention to improving their communications capabilities without adding crushing costs. A laser might, as NASA demonstrated earlier this year, be able to hit a gigabit per second – but the kit's expensive. So a pair of researchers from Florida-based …
Very nice! I've taken to paying attention to various NASA channels. It's surprising that many of the various projects live right now don't use all that much bandwidth. True, if you want all the data immediately you need a stupendous link. Most things don't though. Anyway, worth a pint or three.
The PV panels on a satellite accept a wide range of light frequencies and have a wide acceptance angle so as to get power even if the panel is not perfectly aligned with the sun. This implies that along with the desired signal from the LEDs on the ground there will be a huge amount of background noise from all the artificial lights on Earth (in nighttime) or from the reflection of the Sun's light (in daytime). The resulting signal to noise ratio will be horrible. (Look at pictures of nighttime Earth from space to see the amount of light emitted.) A lens and a narrow band optical filter could improve the S/N ratio by 40dB or more compared to the PV panels.
Yeah, but you'd have to point your lens in the right direction, which is hard.
Picking the AC signal out of the mush is 'just' analogue and processing, no movement. (and blocking the DC from the sun is trivially easy).
I'm impressed that they got this much bandwidth, though. Good stuff.
Stick an aerial on an oscilloscope and see the noise present.
Pass the signal through a narrow bandpass filter and you'll see an actual signal.
Do the same thing but with the signal coming from the PV panels.
If you're looking at a particular modulation frequency, you can pick out the right signal even if there's a lot of noise elsewhere.
This post has been deleted by its author
Ham radio operators have been operating digital transmissions at (sometimes below) the noise floor for years albeit mostly at slower speeds. Various modulation and encoding schemes allow differing combinations of SNR requirements and data rates. What these guys are doing sounds like some variation and improvement on MFSK-16, DominoEX, Olivia or similar. Those date from the early 2000's and there are likely newer modes with even more capability, I just haven't kept up.
But saw Ackerman on the telly tonight, so can't be the really good stuff ...
"Aphelion Orbitals design uses a simple QAM-16 modulation scheme to make it easier to correct atmospheric distortion in the signal."
As for AO, that one sentence explains all of their FSO expertise.
Been told drinking causes brain damage, or did I remember that wrong ...
A slightly unfair comparison. To be fair you would need a system with thousands of satellites. The other down side of using light is how will your weather satellite tell you if it is cloudy?
Looks like a big heatsink on the top and a bigger heatsink on the bottom, with the copper heat pipes leading away from the high-power LED into the bottom heatsink. My guess is that the bottom heatsink at least is for ground testing as the primary method of rejecting heat in a vacuum is radiation, and having a lot of parallel surfaces very close to each other is less than optimal.
The LEDs are only used for the downlink. (A 140W laser is used for the uplink.) They transmit using 1000 lumens worth of LEDs with a peak power of 140W and use lenses to focus the beam. It is reckoned it will emit a 19.9 dBW signal -- that's visual magnitude of 1, so naked eye visible under ideal conditions.
Those fins are a conventional aluminium heat sink. But the LEDs only transmit for 2min/orbit and then they let the heat dissipate. Radiation benefits from surface area every bit as much as convection, so perhaps.
The signal is received using a commercial 30cm telescope. Accounting for losses that's a -107.3 dBW signal giving a speed of 8.85E4 bit/s or 700 photons per bit. But they plan to use "deep learning" on both ends to develop a modulation scheme -- they haven't done so,.
They don't say much about the uplink to solar panels, referring to other work. But it uses a laser mounted on the receiving telescope.
"...reduce the accuracy you need to aim a signal at the satellite, by using a large photon-collection surface that's already common in space, the satellites' photovoltaic (PV) panels, as the receiver."
You'd only believe this claim if you image that a laser beam aimed from Earth to a satellite arriving at the satellite as a tiny red dot; the system struggling to maintain precise aim on the tiny phototransistor on the satellite.
If you understand real world dispersion of laser beams (small, but vastly non-zero), then you'll understand that the mythical phototransistor is *effectively* the same size as the PV array. Because the beam width of the laser beam is hugely vast in comparison, over such distances.
Especially on a tiny cubesat. Less so on the ISS, where this technique is not applicable anyway.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
