I need to transmit small amounts of data (~5kb/s) in the fastest way possible over the distance around 2500 km.

  1. What is the the best way to carry out such transmission?
  2. How could I estimate or simulate the time of signal propagation? Some program packages maybe?



Statement about 5kb\s might be confusing, I'd like to clarify:

Most of the time the data will be contained of Zeros, and only rarely there will be Ones. In other words nothing happens most of the time, but when something's happened – I need to know as fast as possible, faster than by the Internet (which in addition won't be always available).

So, the question is about the complex topic of HF propagation. I need to operate mostly in the daytime.

  • 1
    $\begingroup$ From where to where? If North America, Europe, Asia, I suggest you use the Internet especially to fit your "I need to transmit..." part of your question. When you need to do something, pick the easiest most practical solution to do the job. If land to a ship at sea, use a Sat Phone to send a digital signal over an established link through the Sat Phone signal (a guy I know actually does this). I answered part 1. To answer part 2, just do tests and measure all the parameters you want. $\endgroup$
    – K7PEH
    Dec 17 '16 at 5:44
  • $\begingroup$ Thank you. I guess the initial question was not well stated, please take a look at the update. $\endgroup$
    – Kamerer
    Dec 17 '16 at 14:51
  • $\begingroup$ Are you sure then you want to transmit that much (5kb/s)? This sounds a lot like your actual information rate is much lower $\endgroup$ Dec 17 '16 at 15:56
  • $\begingroup$ I am curious about the purpose of your need. It does not sound like amateur radio where each end point is a licensed amateur radio station. Is this for some kind of experiment or a commercial (aka business) function? By experiment, I am think of something like the Italian OPERA neutrino experiment where they thought neutrinos were traveling faster than light but the error that destroyed that idea had to do with the times being measured by fiber optic cable. The distance was 730 km. $\endgroup$
    – K7PEH
    Dec 17 '16 at 16:48
  • $\begingroup$ The time for light to cover 2500 km is approximately 8.3 ms. However, you will be sending over a larger distance so the time will be more than 8.3 ms. You have not stated your minimum threshold of transit and you have not stipulated the kind of communications protocol you are using as for some protocols, more than one transit will be necessary (e.g. request for transmission of data and then wait for receipt is at least two passes over the path). $\endgroup$
    – K7PEH
    Dec 17 '16 at 17:00

This answer applies to a previous, broader version of the question and remains in case it is of more general interest.

A distance of 2500 km rules out a line-of-sight link due to the curvature of the earth; you would need a tower a hundred times taller than the tallest manmade structure. Therefore, in order to achieve radio communication, you must reflect your signal off something, or have at least one intermediate station which receives and retransmits it.

In any case, the latency is essentially the length of the path divided by the speed of light.

Some options for pure radio communications, roughly in order of increasing latency but not necessarily practical, are:

  • HF radio. Here, your signals are reflected/refracted by the ionosphere, so that they can proceed around the curvature of the earth entirely on their own.

    The catch is that the ionosphere is not always in a condition to serve this function — primarily dependent on the time of day, but in general as variable as the weather. HF propagation is a quite complex topic.

    Also, a good HF station requires physically large antennas. They can be as simple as wire strung between two trees, though.

  • Relay through a satellite. This increases latency because the signal is taking a longer path (two straight lines meeting at the satellite rather than a “great circle” path).

    The time taken will depend on the satellite's orbit, which also determines whether and when you can use it; notably, a geosynchronous satellite is apparently fixed in the sky so you can use it continuously, but is so far out that you will always have a quarter-second of latency.

    You need a satellite to use. All (nearly all? I forget) amateur radio satellites are in low earth orbit (so moving rapidly over the ground and not covering a very large area), and generally heavily used the easier they are to contact. Commercial communications satellites are available but would require some contract to make use of.

    You need a ground station, that is, a radio station that is capable of making itself heard by the satellite. This means directional, steerable antennas, for low orbit satellites; geosynchronous ones still need a directional antenna but it can be installed in a fixed direction.

  • Relay stations on the ground along your path.

    Here, your problem is finding land you can put up towers on, or using a network someone else has already built. Also, the relays will add latency — more or less depending on whether they are full-duplex and how efficient they are.

    The amateur APRS network is in theory capable of supporting this, but in practice rarely gets a packet a long distance on purely RF relays and uses the Internet as a fallback.

  • Reflect your signal off the Moon!

    This is much like using a low-orbit satellite, in that you need high-gain (large) steerable antennas, and it adds 2.5 seconds of latency because the Moon is even farther away.

There's a good chance that your best option is to take none of the above and instead use the Internet, which will do a really good job of getting your data where you need it fast — if you can find Internet access at both ends. Or you could use a radio link over a short distance at one or both ends and transfer the data over the Internet the rest of the way.

  • $\begingroup$ Thank you for such a thorough answer! Please take a look at the question update I made. $\endgroup$
    – Kamerer
    Dec 17 '16 at 14:49
  • $\begingroup$ @Kamerer Thanks for clarifying. I don't have a specific answer to your revised question. $\endgroup$
    – Kevin Reid AG6YO
    Dec 17 '16 at 14:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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