Cave-Link: Underground text communication system | Hacker News Reader

Cave-Link: Underground text communication system

cavelink.com

203 pointsL_22696 comments

> Because all transmissions are secured by checksums and automatic query, no transmission errors are possible.

Just a reminder to people implementing systems with error detection: undetected errors are always possible. Checksums may be just fine for this application but if you need to achieve some target error rate you may have to consider error detecting or correcting codes that fit with the interference you find in your transmission channel.

[edit] Error detecting checks work not by making errors impossible but by making them unlikely. It is part of the work to quantify your design to show that probability is low enough for your goal.

If you hashed every 1000 symbols with a 512 bit hash to check message integrity and retransmit, a hash collision would be practically impossible, so "no transmission errors are possible" is perfectly fair to say under some circumstances.

Also possible they are using a 1 byte CRC heh

A piece of communications gadgetry even mentioning checksums is like a car salesman proudly claiming "... and this baby comes with working brakes!" The very pronouncement leads to worry, and additional questions (like this very thread, QED :-) ).

My inexpert digging came up with: They use APRS packets, which use AX.25, whose framing includes a 16-bit Frame Check Sequence, which looks like it came from HDLC, and is a 16-bit CRC-CCITT. Phew.

Is this what they are doing? Seems unlikely that you would transmit so much overhead when the data rate is so low, unless you needed to overcome a lot of noise.

It’s about the numbers, there’s trade-offs for each specific application. I would encourage people to do the math and see if their design makes sense for their goals.

There's virtually no benefit to dedicating so much of your message to checksums with any remotely efficient algorithm. More practically, any CRC >= 32 bits is probably overkill and any CRC > 64 bits is definitely overkill until you get up into gargantuan message sizes.

Agreed, for this application (texting) a simple approach is probably best. You have to look at the particulars of your domain, it’s not one size fits all.

Sounds like you don't understand secure communications and basic concepts like bit-error-rate, QoS and reliability.

Hash collisions have NOTHING to do with effectiveness or reliability of communication. A feature is a benefit!

Sounds like you simply didn't understand my point.

If you're concerned about one or two bits being toggled, a 512-bit hash will merely increase the chance of bit errors; on a noisy medium, using such a big "checksum" will BOTH detect and cause a high number of packet faults.

The German version is more clear, "automatic query" is translated more literally as "automatic callback", by which I assume is meant that the device requests retransmission in case of an error.

Yes, if an error is detected there’s an automatic retry. I just want to remind others that checksums or any scheme has undetected errors, in which case a retry doesn’t happen. The human in the loop will say, that was garbled, could you repeat?

Again this design may be totally fine for this application. I am bringing this up for other engineers because people tend to hand-wave this away.

"Only a Sith deals in absolutes."

that line (along with enough experience) helped erase words such as "always", "never", and "impossible" from my technical vocabulary :)

I started to reintroduce them into my vocabulary. I found myself qualifying nearly everything, and it decreases the clarity of the discussion, feels lawyer-like. I'm not drafting a contract (most of the time), but trying to communicate a complicated principle in a few words, where every extra word detracts from the meaning.

True story: an embedded target that was loaded over the network (I think TFTP) would boot slowly and then crash when loaded from Workstation 1, but not Workstation 2.

Turns out there was a bad port on the Ethernet hub, but 1/65536 corrupted packets would get through because of the 16-bit checksum.

Go-lay is arguably the best forward error correction for up to 24-bit.

https://en.wikipedia.org/wiki/Binary_Golay_code

Extremely / Very Low frequency radio is fascinating and is surprisingly not used that often. Radio propagation characteristics include being able to pass through water and ground, and continually bounce across the ionosphere / circle the globe. The navy experimented with this for some time to talk to submarines but seems to have ended the experiment.

Additionally I learned that lightning strikes generate a high burst of low frequency radio and commercial lightning detectors (which I use while hiking) actually is just a radio receiver on a harmonic of the lightning signal.

I'm a noob regarding radio: do I understand it correctly that low frequency radio signals can more easily be used with more solid carriers while higher frequencies allow more/faster bandwidth? (I'm totally oversimplifying here)

Yes

You need bandwidth to transfer more data (there is a theoretical limit of how much data you can transfer at a given bandwidth with a given signal to noise ratio).

The lower you go with frequencies, the larger antennas you need (quarter wavelength is usually the minimum... for FM radio (around 100MHz) it's 75cm which is very realistic length for a car or a house radio... at 14mhz (ham radio freq), it's 5.3 meters, which is still doable if you live in a house, and at 1.8MHz (another ham band) is ~40 meters, which is hard.. going lower is even harder, and special "tricks" are needed.

The second problem is the bandwith itself... some wifi equipment supports bands up to 100MHz of width (to simplify, you use all the frequences from eg. 5.1GHz to 5.2GHz), and if you want to stay at low frequencies, it's impossible to create such a wide band, because the antenna sizes are so different at 1MHz and 101MHz, the propagation is different, etc., so usually you're stuck with narrower bandwiths at lower frequencies (and of course, many of the low frequencies are already in use.. eg for fm radio).

But generally yes, higher the frequency, more directly (in a straight line) it goes, and you get more losses in stuff between the transmitter and receiver.

Can the antennas be coiled to achieve a long length while not being strung , for example, 40 meters? I recall car antennas that used to have coils long ago.

There are coil antennas, but they work magnetically. The 40m straight wire works electrically. The key difference is that the electric antenna interacts with more of the wave, so it's stronger in the far-field, while the magnetic antenna interacts more localized. The latter is why/how NFC works at 13 MHz (23 m wavelength) with an antenna that fits on a credit card.

Not really, atleast not the classic designs and if you want to transmit. You can bent them a bit, but the total lenght has to be around the quarter wavelength

Old school 27MHz CB radios would have 1/4 wave antennas 2.75m tall. In practice, for vehicle mounted use, they'd almost always have bottom or centre coils to allow overall lengths down in the 1.5-2m sort of range, with compromises accepted there for practicalities and convenience. Fixed base stations would often use a "proper" 2.75m antenna though.

Thanks!

Generally speaking, yes. With RF there are a handful of encoding methods, but in very simple terms binary 1's and 0's correspond to the high/low peaks of the frequency. Higher frequencies can therefore transmit data faster because you have more "peaks", and opportunities to encode data per interval of time.

Higher frequencies have less power in terms of ability to travel a great distance or through materials. It is very much like audible sounds, you can hear bass notes from someone playing music far away, but you'll lose the highs and then the mids after a relatively short distance.

One of the many tradeoffs in RF networks is trading range for throughput.

> but in very simple terms binary 1's and 0's correspond to the high/low peaks of the frequency

This is AM, "Amplitude Modulation"

Wiki has some good articles on other modulation techniques https://en.wikipedia.org/wiki/Amplitude_modulation

Oh the comparison with audible sounds is brilliant, thank you.

Bandwidth, literally the width of the band. Don’t think of it as transmitting at one frequency but a range between two, a band.

Information rate depends on the width of the band. For example wifi uses 20, 40, or 80 MHz bands. Easy when your signal is around 2.4 or 5 GHz.

But you can’t do that at 100 Hz or 100 kHz or 100 MHz because… well the numbers aren’t big enough. You could have a funny radio that transmitted from 1 to 80 MHz, but you wouldn’t say your signal was “at” 100kHz.

That make sense?

Yes thank you!

Lighting strikes also feed the very cool https://en.wikipedia.org/wiki/Schumann_resonances

I'm reasonably sure VLF and ELF are very much in use but it seems as though ELF was abandoned by the US Navy. The Navy does continue to use VLF though.

https://www.thedrive.com/the-war-zone/25728/chinas-new-york-...

http://www.nukeresister.org/static/nr135/135elfcloses.html

There's a huge antenna in Colorado, I think managed by the Denver Air Route Traffic Control Center which is located in Longmont, that's used as a backup for transatlantic flights communications.

How huge? Aviation uses HF (shortwaves) and not VLF as a backup for translatlantic flights (see for example https://thepointsguy.com/guide/how-pilots-communicate-with-a...). HF antennas can already look huge ;-)

Trans-oceanic flights go far beyond the VHF radio horizon to land so routinely use HF radio. It's not a backup, it's the norm.

Example: https://www.iaa.ie/air-traffic-management/north-atlantic-com...

"Huge" is probably not that big vs VLF. Maybe on the order of ~50m width (~3MHz / ~100m wavelength for ~50m half-wave antennas). VLF is more like wires across entire valleys - kilometres.

At the frequencies and distances that these cave "radio" systems are operated they act more like two coils of wire with a mutual inductance in the near field (<1/4 wavelength away). This isn't propagating electromagnetic waves (radio), at least not much or with any fraction of efficiency.

I never understood why visible light goes through water but almost every other frequency is blocked?

Why? The properties of water were optimized to make life possible!

I think this reasoning may be right, if expressed backward. "Visible light" is just that which life evolved to care most about, because it was what was available in the (shallowish) water.

in the case of the whales, whom returned to the deep sea 40-50mya and have since rarely had a use for visible light, it's interesting to wonder what they have evolved to care most about if no longer visible light, especially in the sense of their song, and when you compare the visual acuity of their cousins the dolphins in the same timespan, and the incredible hyper-evolution of the human eye since ~1mya

I just meant the physics of it.

The properties of vision were optimized to make seeing through water possible!

The issue is even more complicated. The absorbtion at the red end of the visual spectrum is two magnitudes larger than at the blue end. This is why water appears often blue, even against a white background.

Wikipedia has some more details: https://en.wikipedia.org/wiki/Electromagnetic_absorption_by_...

I just need an ELI5 of that Wikipedia page and I’m all set. Good find though.

This reminds me of the time I went looking for a satisfactory answer for why the sky is blue.

It turns out a surprising number of people have a surprising number of different opinions and explanations on this. You pretty much need a PhD in physics to fully grok most of them. And every time you think you've read something that at least _sounds_ concrete, someone else comes along and says, "yes and no, the _real_ mechanism is this," ad infinitum until you wind up on the fringes of scientific knowledge.

In the end, hours of research later, as best as I could tell, it came down to: the sky is blue because air is blue. But it's a very faint blue, so you can't actually discern the blue until you look through a _lot_ of air, for example in the sky.

> the sky is blue because air is blue.

Things have color because they scatter/reflect/refract/absorb/emit differently for different colors of light.

Air looks blue because it scatters blue light more than other colors. The blue components of sunlight reach you from all parts of the sky, while the non-blue components reach you mostly from the direction of the sun. At sunset, sunlight from the direction of the sun passes through much more atmosphere than during the day and most of its blue light scatters away, leaving red and yellow. [0] If air didn't scatter light, the sky would be black like it is on the moon.

Air is a gas. Gasses do not reflect light.

Air refracts sunlight and makes the sun appear flattened at the horizon [1].

Air absorbs most non-visible colors (UV and IR) [2]. Animals evolved eyes to see only the colors of available light.

During aurora, the air emits green light.

[0] https://en.wikipedia.org/wiki/Rayleigh_scattering

[1] https://en.wikipedia.org/wiki/Atmosphere_of_Earth#Refractive...

[2] http://funnel.sfsu.edu/courses/metr104/F13/summaries/Absorpt...

If there were a open band somewhere else, that is what we would have evolved to see.

Is just the dynamics of how light interacts with the atom and its bonds.

It's the other way around. Our eyes evolved to be sensitive to (some of) the wavelengths to which air and water are transparent.

There's nothing innately special about visible light. Just about every substance is transparent to some wavelength or other.

Because people haven't really explained the specific mechanisms:

You can think of a material as a bunch of simple harmonic oscillators, which are driven by an applied field (electric or magnetic). These oscillators have different natural frequencies, and so couple to different wavelengths. For instance, if you play a loud note next to a piano, you can see the corresponding piano string start to vibrate, but the others may not.

The interaction between these oscillators and the wave are what lead to the dielectric constant (and therefore the slowing of waves), as well as absorption (which can be thought of as the imaginary component of the dielectric constant).

In a real material, these "oscillators" are really any method of energy storage that can couple to the motion of charges (i.e. current). These include, but are not limited to: - rotations (in a gas or liquid) - vibrations (in any state) - electronic transitions - electronic movement (in the case of a metal) - displacement (in any state)

In a single molecule, many of these mechanisms would have discrete natural frequencies. But in a solid or liquid, interactions lead to a continuous band structure (especially for things like vibrations).

For water specifically, the below visible range is quickly absorbed by vibrational and rotational energy modes, while the high end of the UV range is absorbed by electronic transitions. Other materials have similar sweet spots for transmission, but at different frequencies. For instance, materials like indium tin oxide (ITO) are designed to be conductive, but not at the high frequencies of visible light, making them transparent. As another example, metals are reflective below their plasma frequency (related to the speed the 'electron sea' can move at), and transparent above (X-rays operate in this region of transparency).

If you want more information, I can recommend "Optical Properties of Solids" by Mark Fox.

Russians still use extremely low frequencies for communication with submarines (an unidirectional communication system called "Zevs"). Transmitter is located on the Kola Peninsula, frequency is 82 Hz, bit rate is a few bits per minute, power consumption is in the multi-megawatt range. Waveguide is formed by the surface of the earth and the ionosphere.

I'd like some intuition --- is that very low frequency? Is that very high power? Can I have some context for how I should think about 82 hz and multi megawatt? (For example, I know that few bits a minute = well, that's a couple ASCII characters a minute)

82 Hz implies a wavelength of ~4000km in vacuum. It’s extremely low frequency by comparison with the kind of radio spectrum we usually use.

The Wikipedia page gives a good overview: https://en.wikipedia.org/wiki/Radio_spectrum

Isn’t that audible? And with megawatts of power… think of the whales!

Its still electromagnetic waves not pressure waves so they're not audible.

Still, I'd recommend not being very close to the transmitter if its putting out megawatts of power.

Nah, it’s electromagnetic & your ears are too small to intercept the transmission to any kind of significant extent.

A few MW sounds a lot, but the inverse square law applies - once you get to the ocean the power density is going to be very low. Plus penetration into the ocean itself isn’t going to be great - there’s an impedence mismatch between the water and the air.

You know that hum you get when mains powered audio get gets a certain kind of interference? That's 60Hz (or 50Hz depending on where you are) of moving air. It's not the 50Hz electromagnetism you h4ear though, it's that getting picked up by the amplifier and played out through the speakers which move the air.

87Hz is right about the second lowest F key on a piano. (the 21 key up from the lowest note on an 88key piano), or the low E string on a guitar at the first F fret.

Here are some names we give to radio bands. These names are somewhat dated, as even “high” frequency is pretty low.

Low Frequency (LF) (30 kHz to 300 kHz)

Medium Frequency (MF) (300 kHz to 3 MHz)

High Frequency (HF) (3 MHz to 30 MHz)

Very High Frequency (VHF) (30 to 300 MHz)

Ultra High Frequency (UHF) (300 MHz to 3 GHz)

HF is used by amateur radio operators to communicate around the world, and has sufficient bandwidth to carry voice or low speed data (300 bits per second). Commercial airliners use HF over the oceans too. Probably some military stuff too.

Most other radios on the planet is operating on a higher frequency, in VHF/UHF or higher.

As for power, 100 watts is enough to send an HF signal around the globe. Commercial AM radio stations might operate up to 50,000 watts, so a megawatt is a lot more.

If you don't already know just how low 82 Hz is - then the lowest frequency band you're probably familiar with is AM radio. Which uses around 1,000,000 Hz.

And as others have mentioned, yes, megawatts is a lot. That's the same order of magnitude as what most power plants produce.

And your power points in your house use 50 to 60Hz, depending on where you're from - and almost nobody even considers that "radio", even though technically it is.

Typically antennas for radios are 1/4 wavelength of the transmission frequency long (for various reasons but mostly related to transmission efficiency).

So as comparison, the 2.4GHz WiFi/Bluetooth radios in your phone have a 125mm wavelength so probably have 32mm long antennas (most likely folded in interesting ways to make then physically shorter). Old school 27MHz CB radios have an 11m wavelength, so ~2.7m antennas (in practice, at least on car mounted antennas, they'd coil some of that length up at the base out in the middle of they antenna, to keep the size down to 5 or 6 feet.) An 87Hz radio has a 3,500km long wavelength, so would want (for peak efficiency) an antenna almost 1000km long.

On top of that, power requirements to get a certain range go up with the square of the distance. Your phone WiFi probably maxes out at 200mW at it's 125mm wavelength and gets maybe 200m or less of max range. It's about 20,000km to get halfway around the planet, so ignoring a bunch of other important things, that's 100,000 times as far you'd need to transmit with 10 billion times as much power to get the same signal strength at that distance. That'd need ~2GW. Your phone battery wouldn't;t last long trying to do that...

(There's another complication abiyut the power requirement changing with wavelength, but I'm not quite enough of a radio geek to know that and the numbers I quickly googled up doin;'t make much sense, so...)

82Hz is an E2, the 19th white key from the lowest on a piano. [1]

So it's fairly low but completely audible.

For megawatts, imagine a million bookshelf speakers playing the lowest note they can play, or 10,000 subwoofers playing the highest note they can play. Most people listen to music at less than a Watt power.

I certainly wouldn't want to live next to that transmitter.

[1] https://en.wikipedia.org/wiki/Piano_key_frequencies

You're fine with hojillions of neutrinos zipping through you. It's not just about power, it's about interaction. How's living near that 82Hz transmitter compare to constantly living around electrical wires acting as 60Hz transmitters? Or a substation, for that matter?

It's much worse. That's audible. 60Hz wires don't shake the whole house.

The US Navy still uses VLF, though ELF comms seem to be retired at this point.

https://en.wikipedia.org/wiki/Communication_with_submarines#...

> surprisingly not used that often.

VLF (and the lower frequency parts of the radio spectrum) aren't used much anymore because power-efficient antennas are too large to be practical. If RF energy enters an antenna that doesn't resonate at the correct frequency, much of it will be reflected back and forth within the antenna until it all dissipates as heat due to internal resistance.

A simple kind of antenna is a vertical antenna. These need to be 1/4 the wavelength of the RF signal. You can get by with more or less than 1/4 wavelength, but you'll waste increasingly more energy.

Wavelength is inversely proportional to frwquency. If you've got a microwave signal with a wavelength of 30cm, that's practical. VLF frequencies go up to 30 kilohertz at the high end, which is about 1 kilometer (!).

So if you want to transmit at decent power, you need a big antenna and a big power source, because you're still not going to have a perfect match and some will be wasted. This is what a VLF antenna array looks like:

https://en.wikipedia.org/wiki/VLF_Transmitter_Cutler

Notice it's absolutely enormous.

Back when we didn't have high frequency electronics, commerical communication was done in the hundred kilohertz range. You can see here what those coast station antennas looked like:

https://www.radiomarine.org/historic-coast-stations

This was used in the recent rescue in Wales [0]. I seem to recall a colleague telling me years ago (in about 2000) that he got stuck in a cave once (somewhere in South Wales I think) due to unexpected rain causing flooding and there was a phone available that could be used to request rescue (or to tell people you were going to sit it out).

[0] https://www.bbc.co.uk/news/uk-wales-59219380

This site is a bit scarce on details. There's more details and some photos here: https://de.wikipedia.org/wiki/Cave-Link http://www.hoehlenverein-blaubeuren.de/index.php/cavelink-te... (Use Google Translate)

I wonder how the licensing issue referenced in the Wikipedia article is handled in the UK. I imagine Ofcom will have had to approve and license these installations, or will they have made these exempt (I can't find any evidence of this)?

as far as I know similar general licensing exists for such purposes. Lots of inductive applications need it.

I can't see these devices specifically mentioned in https://www.ofcom.org.uk/__data/assets/pdf_file/0028/84970/i..., though the frequency range they operate in (20-140kHz) is allowed to be used by induction loop type applications, at a suitably low power.

... but the cave link uses relatively high power levels (20-30 Watts if I remember correctly)

Inductive stove tops are also In the KW range

I'm really surprised ELF are not commonly used for this. Even a simple single-frequency CW transmitter for emergencies.

I wonder if audio frequencies below 20KHz can be used to make a cheaper system.

At the extremity of that band, fewer nations have constructed transmitters than have nuclear weaponry.

You can combine acoustic with radio in TARF (yes, really).

I meant acoustic frequencies in ELF, not using sound... but TARF looks amazing.

see e.g. https://www.mit.edu/~fadel/papers/TARF-paper.pdf for those who didn't hear about TARF (translational acoustic-RF communication)

Something that is mentioned here and not on the main page: the devices form a mesh network and can relay messages between each other.

Seems strange, feels like a pretty big selling point.

See also the HeyPhone and Nicola systems built by underground boffins in the UK and France:

https://www.shropshirecmc.org.uk/radio.html

http://www.scavalon.be/avalonuk/technical/radio1.htm

https://www.electronicsweekly.com/blogs/engineer-in-wonderla...

https://www.caverescue.org.uk/nicolaradio/

http://site2.caves.org.uk/radio/comms_in_caves.html

Similar tech was used during the rescue of the boys out of a cave in Thailand back in 2018

https://bcra.org.uk/creg/heyphone/ https://hackaday.com/tag/heyphone/

This uses VLF RF, but there was another interesting thing I have read about long ago called "earth mode" or "conduction mode" maybe. You put two conductors in the ground some distance apart (10s or hundreds of feet) and inject the signal into the ground. The signal causes a current in the ground which will cause voltage differentials that can be picked up miles away with a similar receiver. I believe it is also low bandwidth, but I often wonder if you could create a mesh network of these. Also would the FCC actually be able to regulate this since it is not actually a radio wave?

I think it is still a radio wave, but propagated through soil and moisture rather than air.

Maybe, but I thought it was different since it uses conduction. For example I am pretty sure you could detect a DC signal, whereas with radio you cannot.

That's pretty cool!

[V|E]LF stuff is used for many things. For example, the Navy uses it to communicate with subs.

Reminds me of an interesting Great War technology. During WW1 the British used a system called the "power buzzer" which used two spaced electrodes and sent an alternating current through the earth. Apparently this could be picked up and amplified by another ground electrode at some distance, and was used to send morse code communications. I'm struggling to find a good modern explanation but here is a scan of the original military manual:

https://cgsc.contentdm.oclc.org/digital/collection/p4013coll...

Also in caving news this week https://www.bbc.co.uk/news/uk-wales-59203695

Yeh, this is the system they used to communicate during the rescue.

Could see this being used in the Mammoth Caves National Park. A lot of the cave is undiscovered and having reliable comms might make that work a bit safer.

If you liked this you should check out the British Cave Research Association's Cave Radio and Electronics Group at https://bcra.org.uk/creg/

A quantum entangled communication system wouldn't have any 'transmission' issues (through mediums, etc), correct?

You can’t transmit information with quantum entanglement.

Really? But can't you influence 1 side of an entangled pair, then measure the other side, and you can pass information through the correlation? I thought quantum-entangled FTL communication devices were a legitimate thing.

Nope. You can’t change the outcome, you can only make an observation on one and then and only then know the state of the other. What is being observed is entirely random.

Information cannot be transceived faster than light. The thought experiments and whatabouts and whatifs are fun to discuss, especially after drinks - but seeing which spin or collapse or whatever isn't information in the actual definition of information.

Information involves entropy and all that stuff is bound by the speed limit.

How do the Mole People feel about this?

Suppose you're lost inside a cave... In a great feat of luck, you find a cave-link base unit. Nice let's ask for help! You then discover the system is in German and you can't understand how to use it: https://www.cavelink.com/cl3x_neu/index.php/en/component-ove...

Ok, I know it supports other languages, but there's no reason for the main site not to use English to illustrate it. Or better: make the interface intuitive enough that you can use even without reading instructions. We did it with smartphones.

These are highly specialised pieces of technical equipment with a tiny potential market. The use case isn't fixed unattended public installations to be used by untrained randoms, but ad-hoc use by trained users like cave rescue teams and speleologists. As for the UI, production (like much in the caving world) is semi-commercial at best; an iPhone it isn't and frankly it doesn't need to be.

Reply on news.ycombinator.com