Find myself wishing such tech was prevalent during the early days of mobile phones and massive external aerials, but that's just the inner Jedi in me.
> during the early days of mobile phones
Probably not a good idea, in its original form, it requires a lot of power from the mains to maintain its arc discharge, which is what turns it into an antenna, kind of like a vacuum tube that needs a constant heating power. But Wikipedia says antenna-on-chip is possible, which is quite interesting.
An arc discharge is a hot, high current density process.  The discharge in a fluorescent lamp is a glow discharge: low current density, moderate temperature and extended volume. 
I wonder if something like this could already be in use, since it becomes invisible to radar as soon as the plasma generator is turned off?
There are probably ways to filter for specific RF frequencies on the cover so only those in specific bands can get through, then the array itself might absorb some portion of the energy that gets through. The array also might be at an angle so waves are reflected in a different direction, or there might be radar absorbing material placed in specific spots that helps minimize reflections. I think the arrangement of elements in the phased array itself could also lead to reduced RCS at certain angles. Again these are just educated guesses, I would imagine the details are classified.
As it turns out someone already patented "Antenna of ionized air": https://patents.google.com/patent/US2760055 The patent expired in 1973.
Meteors punch ion trails through the atmosphere that serve as antennas for various purposes.
Is there a layman's explanation about what the advantages / why there are advantages here?
It talks about turning off the plasma antenna, is that any different than simply not using a regular antenna? ... or what they mean by stealth or "resistance to electronic warfare and cyber attack".
When the plasma antenna is turned off, it stops existing. There's nothing there to attack or detect.
A possible future advantage would be shaping the plasma with magnetic fields, to create customizable antenna geometries, or geometries that electrically extend beyond the physical bounds of the equipment. Imagine using a laser small enough to fit in a backpack to create a 1/4 wave antenna long enough to transmit or receive at 30 kHz, with length 2.5 km .
You could create a plasma antenna 2500m long, send a message to a submarine 150 m below the surface, then it just disappears when you turn it off.
It would be unlikely to glow brightly enough in the visual spectrum to see in the daytime, but it would be pretty cool at night.
This, however lets you dynamically choose the antenna length, and each frequency has a perfect length that works best for it.
So the two are complimentary, but at different parts of the process.
Did you even read the article? They mention Solid State Plasma antennas, which can be fabricated using standard silicon chip fabrication techniques.
For security applications, the ability to turn it off is kind of nice; no antenna = no backscatter. OTOH how stealthy can a plasma antenna be when it is on?
What gives, the Web? Donate to archive.org today!
Presentation from Igor Alexeff and Theodore Anderson (University of Tennessee):
> At satellite frequencies, they exhibit much less thermal noise and are capable of faster data rates.
If you look at the thermal analysis, it is incomplete. They are only considering the thermal noise of the reflector. The noise temperature of an antenna is a function of temperature AND aperture efficiency, with the latter set by the conductivity of the reflector. Plasma is not a great conductor (compared to metal), as is shown by the nested antennas. High performance satcom (e.g. NASA DSN) not only cool the electronics, they cool the feeds also.
There is another paper where they show a tube covering a LNB. Looks like some BS. I don’t see any actual measurements of noise temperature (which are very easy to do).
I didn’t see any of the papers published in IEEE APS.
I design antennas for a living, and this sets my BS detector off, and it did 10 years ago too. They look fine for TX (the patterns look good), but the noise data is conspicuously absent.
Here is another noise analysis, but again it is incomplete. Just show some measurements.
You don't need high temperature for plasma; just low pressure. There is a big difference between glow discharge and arc discharge.
They just need to hook an antenna to a spectrum analyzer (with low noise path) and measure the noise density. If it is anywhere slightly above -174 dBm/Hz, then it is not going to function well as a satcom antenna.
That picture they show of the tube taped onto the LNB feed, they say it “intercepted” the signal. Do they mean it blocked it? Sure, it will if the plasma is conductive enough, but it should also drop the noise coming out of the receiver. Maybe it swamped the receiver with noise. Can’t tell. If they can pattern the lower frequency antennas, they can certainly take the time to do some noise measurements.
Maybe the issue is exciting the plasma, and that is noisy. I’d think you couldn’t do it with pulse excitation unless you limited the rise time. You could do it with CW, say a magnetron, way outside your operating band.
Did not know you could make an antenna out of a plasma; this is fascinating; thank you for expanding what I thought was possible!
I have often toyed with the idea of making oscilloscopes, and spectrum analyzers with them, but that would take a long time to explain...
Also, I miswrote it's Glow Discharge Detector
Another idea is repurpousing plasma display panels as a grid of GDD detectors, as an imaging array.
For an example circuit using a neon indicator lamp see figure 1 of the 2010 article:
This 2016 article measures the light emitted from the neon indicator lamp instead (with FMCW scheme to detect distance as well)
This setup improves response time (bandwidth) of the output signal.