18
\$\begingroup\$

So I know thermo cameras are sold. Police/Fire use them all the time. You point your little hand held black box at a bush and you can see if there is a creature living in there or at a burn zone and see where the hot spots are.

Is there a device that will do the same for RF signals? VHF up to 2.4GHz I was thinking it would be a great aid in locating RF noise in a sensitive environment.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ It's called radio telescope. \$\endgroup\$
    – Peter G.
    Commented Oct 2, 2013 at 12:44
  • 2
    \$\begingroup\$ Or a passive phased-array radar. \$\endgroup\$ Commented Oct 2, 2013 at 16:44

5 Answers 5

19
\$\begingroup\$

So I find it very interesting that all answers until now seem to think in terms of pre-1900s radio technology. In order to productively think about portable or reasonably sized radio imaging techniques, you have to think a bit differently.

The way to receive electromagnetic waves is to produce a material that is opaque and absorbant to the wavelength. Then, the absorbed waves should be converted into an electrical signal to be measured. There are quite a few ways to do this: for instance with visible light, single photons have more than enough energy in them to excite electrons in certain crystallographic structures. So all you need to do is make a relatively conductive bulk material that is opaque to your specific wavelength and all the light of that wavelength hitting the material will have a (significant) chance of generating an electron.

Radio frequencies are a much longer wavelength and by extension have much, much lower energy. The energy and wavelength are an inverse proportional relationship, so like Andy said: 300 million times less energy. This is not nearly enough to excite electrons out of the valence band of atoms, even if you would throw extremely high radiative energy densities at it. Absorbing those photons is no problem, the trick is in how you convert the photons into an electrical signal.

By the way, it is a fallacy that you need a material that is physically larger than the wavelength to absorb it. For instance, water molecules are extremely good at absorbing radio waves, even though they are many orders of magnitude smaller.

The easiest and most intuitive way is to take an antenna that is exactly one wavelength long. This antenna will react purely to the magnetic component of the electromagnetic wave (both of which have the same wavelength), and the antenna will react as a high impedance inductor, creating a current from the magnetic field that is induced. The antenna having exactly the wavelength, it is resonant and will create the largest possible signal from these photons. This is extremely basic physics.

However, you don't need to look at photons as waves all the time. They still also behave like particles, and you are able to 'catch' one even if you have a much, much smaller surface. One way to do this, is to create an antenna on which the incident waves will bounce around a couple times, effectively increasing the path length until it is about the wavelength of the photon. This way you still get the same absorption and resonant magnetic properties of the antenna, but with a much smaller physical size. These are the antennas we use in mobile phones nowadays, colloquially known as 'fractal antennas' (the shape is derived from fractals to maximize path length for all directions of incident radiation).

But this is still not the smallest you can get a detector. It is possible to actively tune a very small piece of absorbant material, and it is possible to make it absorbant in one specific direction. That way only photons emanating from a relatively small solid angle will be absorbed into the detector. This is done with resonance again - a resonant circuit at about the frequency of the light is connected to a conductive radio-opaque material, and when radiation is incident, the resonance point will shift, indicating reception.

This all means that it is not necessary, as many people will think, to have humongous sensors to 'view' radio waves. However, sensors will never be nearly as small as visible light imaging sensors. Even though you can 'cheat' normal optic laws and have smaller viewing angles with smaller optics than you would expect from Airy, the amount of energy in the radiation severely limits how well you can image long wavelengths. You would need extremely long-term exposures, it is definitely not possible to get multiple frames per second. As it stands right now, with the best detector technology we have we're talking about hours or days of exposure with a detector the size of a table, let alone a truly portable radio imaging sensor. Possibly superconducting materials may improve this, but I know of no research in this area.

To get back to you actual question: there is no commercial device that does what you want, yet. There is research in this area though, and it will not be very long until we will have such devices. However, it will also not be long until your cell phone will be able to do RF imaging, with the advent of phased arrays and essentially 'imaging' antennas in phones.

\$\endgroup\$
2
  • 4
    \$\begingroup\$ Everything you said is true. But none of it particularly covers directionality, which is the largest hurdle. \$\endgroup\$ Commented Oct 2, 2013 at 8:15
  • 1
    \$\begingroup\$ @IgnacioVazquez-Abrams: yeah, that is true. I kind of stopped at that point. Directionality is something that can nowadays be improved with either phase difference measurement or anisotropic resonance, but even though you can get better angular resolution that way than a synthetic aperture would give you, you are still looking at multiple degrees of solid angle per 'pixel'. I don't have an answer on how to improve that vector of sensitivity. \$\endgroup\$
    – user36129
    Commented Oct 2, 2013 at 9:12
5
\$\begingroup\$

If you had a bag of sand and spread it evenly on your floor you can draw shapes in it with your finger and make intricate sand castles from it. That's my analogy of visible light. The analogy for VHF/UHF would be grains of sand the size of a football stadium.

Green (the colour) has a wavelength of about 500 nano metres - that's half of one thousandth of a millimetre.

1GHz has a wavelength of about 300mm - 600,000 times bigger.

\$\endgroup\$
3
  • \$\begingroup\$ Isn't imaging achievable with passive sound? Its not necessarily camera-like, but it is done. \$\endgroup\$ Commented Oct 2, 2013 at 10:46
  • \$\begingroup\$ @ScottSeidman: Bats use sonar imaging, which works because the wavelength is ~1 mm. It's not the frequency which matters, but the wavelength. \$\endgroup\$
    – MSalters
    Commented Oct 2, 2013 at 10:54
  • 1
    \$\begingroup\$ Correct, it's the space definition that a small wavelength offers. Consider also a road tunnel - if your car radio is tuned to the AM band, the second you enter the tunnel the music disappears to be replaced by noise and spark plug noise. At VHF, you can get a lot further into the tunnel before the music disappears. AM is about 1MHz which has a wavelength of 300m whereas 100MHz (VHF) has a wavelength of 3m. Bats can "hear" 100kHz sound waves and these have a wavelength of about 4mm. \$\endgroup\$
    – Andy aka
    Commented Oct 2, 2013 at 11:02
3
\$\begingroup\$

The longer the wavelength of the radiation, the larger a sensor you need to detect it. Radio waves, with a wavelength starting in the millimeters, require far too large a sensor to detect in the same manner.

\$\endgroup\$
1
  • 1
    \$\begingroup\$ To be precise: this applies to imaging sensors, which is what you need in a camera. Smaller sensors can detect the radiation - detecting the 50 Hz field of high-voltage lines obviously does not require a 6000 km big sensor - but such small sensors cannot form an image. \$\endgroup\$
    – MSalters
    Commented Oct 2, 2013 at 10:50
1
\$\begingroup\$

This can been done at home by using a directional antenna on a gimbal and an SDR.

It's not portable and not fast, but you can build it yourself and this particular project is open source, so you can basically follow the instructions and get started.

Building a Camera That Can See Wifi | Part 3 SUCCESS!

A group at TUM has also achieved this using radio holography. See their slideshow here (their paper is available online for free: Holography of Wifi Radiation 2016, P. Holl).

Holography of Wi-fi Radiation

It's very interesting work and much faster than the first approach.

\$\endgroup\$
3
  • 1
    \$\begingroup\$ Their rig is insanely slow, and generates an insane amount of data - most of which is discarded. The main problem is their method of capturing power levels. The 8 bit sampling they use means they have to use a lot of averaging to resolve small variations. They do it by collecting several gigabytes of raw data, and post processing. It'd be far more efficient to use a better receiver and a digitizer with more bits per sample. \$\endgroup\$
    – JRE
    Commented Jun 2, 2019 at 8:08
  • \$\begingroup\$ ...and the use of that helical antenna is completely insane. Clearly the authors don't have the faintest idea about antenna design or theory. For pinpointing radio sources you need a parabolic or patch antenna with a very narrow beam-width. (Helical antennas are used for circularly polarized radiation, which is not used for Wifi.) \$\endgroup\$
    – not2qubit
    Commented Jun 2, 2019 at 12:42
  • \$\begingroup\$ It's true, but even with their suboptimal design the resultant data was a clear image of wifi illumination that illustrated the locations of hotspots and reflective surfaces. I'm sure a more sophisticated approach could produce even faster, more detailed, and less noisy results. \$\endgroup\$ Commented Jun 3, 2019 at 18:17
-1
\$\begingroup\$

The way I pictured this is to have enough spectrum analyzers for each pixel. Say you want 1080p resolution, You would use about 1 million spectrum analyzers for each pixel. Than you would need to have 1 million antennas. Granted this would be monstrous in size but would work in theory.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.