I'm interested if it is possible to have a data radio modem that works on the UHF band 400MHz for TCP/IP communication.

What I have in mind is something similar to a router, but using UHF band for more range.

The problem that I come across is that there are so many serial RS232 based data radio modems, but they don't have Ethernet and they operate very slowly. These are common.

Is it that hams are just not keeping up with the technology or is it not possible to modulate a UHF signal to get a few hundred Kbps?

Why not take the technology and modulation schemes used in 3G or 4G cellular networks and apply the same compression/modulation and whatever else into a radio modem that will operate much faster at 400MHz.

3G and 4G networks are extremely fast and 400MHz is only about half the frequency of them, so I'm thinking in theory, you could probably still get some pretty reasonable speed at these frequencies.

Are there any devices out there like this that are not thousands of dollars? Could one be built by an amateur using say a raspberry PI and some type of RF module? Maybe a software defined radio like hack RF?

Are there any legal regulations that talk specifically to this?

  • 1
    $\begingroup$ When asking legal questions, please specify which jurisdiction you are enquiring about. The world assumes "USA" in most cases, but that is most definitely not always the case $\endgroup$
    – Scott Earle
    Jun 21 '18 at 3:22
  • $\begingroup$ It's not SDR, and it's ancient, but here's a project that does something like what you're talking about -- wa4dsy.net/rfmodem.html $\endgroup$
    – Duston
    Jun 29 '18 at 14:11

As you might suspect, the data rate impacts the required RF bandwidth. Besides the data rate, the modulation method, encoding, packet overhead, duplexing, and other factors ultimately determine the required RF bandwidth. As an example, the 4G network has a data speed of ~20 Mbps and consumes a bandwidth from to 5 to 20 MHz.

In the United States, the 70 cm allocation extends from 420 MHz to 450 MHz in ITU region 2 for a total of 30 MHz of bandwidth. For US regions in ITU 1 or 3, this drops to 430 MHz to 440 MHz. In the US, hams on this band may not cause interference to radiolocation services (e.g. PAVE PAWS) that overlap this allocation. This would suggest that the technique would require a guard receiver and be adaptive to lock out bandwidth segments. The 420-430 MHz segment is also allocated on a primary basis to other services and not even available to hams north of "Line A". But the most significant limitation for data in this band is spelled out in 97.307(f)6:

A RTTY, data or multiplexed emission using a specified digital code listed in §97.309(a) of this part may be transmitted. The symbol rate must not exceed 56 kilobauds. A RTTY, data or multiplexed emission using an unspecified digital code under the limitations listed in §97.309(b) of this part also may be transmitted. The authorized bandwidth is 100 kHz.

So, at least in the US, there is insufficient permitted data rate to support your request. Since the US is the largest ham radio market, manufacturers may not have a large enough remaining market to support the development of such equipment even if other regions of the world have sufficient bandwidth allocated to the ham community.

In the US, the the data rate and bandwidth limits are lifted beginning with the 33 cm band (902-928 MHz in ITU 2 only). This band is granted as a secondary allocation and hams must also accept interference from industrial, scientific, and medical (ISM) equipment.

In the 13 cm and 9 cm bands, US hams have been very successful re-purposing equipment designed for consumer WiFi applications to the frequencies allocated to hams in the US on a primary basis where more power and higher gain antennas are permitted. See Hamnet as an example.

Finally, consider that for hams in the US, the TCP/IP traffic cannot contain encrypted or obscured information. This includes all encryption uses of TLS such as HTTPS and WSS type protocols. There is a limited encryption exemption for space telecommand purposes.

  • 1
    $\begingroup$ There's a limit on symbol rate, not bit rate. A "couple hundred kbps" seems technically feasible at that symbol rate with a higher-order modulation. $\endgroup$ Jun 8 '18 at 13:09
  • $\begingroup$ Technical feasibility and real-world feasibility are different things. With fading, noise and multipath, etc, for non-LOS connections, even 56kbps data (or symbol) rate may not be possible. $\endgroup$
    – Duston
    Jun 8 '18 at 21:28
  • $\begingroup$ "Technical Feasibility" means "real-world feasibility", @Duston. Of course you're right, noise and fading limits our data rate. But assuming sufficient CSI, we can equalize, or we can use multicarrier schemes, so the SNR is really our limiting factor: for 56 kb/s, that limit is R=56kHz < 100 kHz· log<sub>2</sub>(1+ SNR) -> 1+SNR = 2<sup>0.56</sup> -> SNR = 2<sup>0.56</sup>-1 = 0.475. So, you'd need worse than 3 dB to technically transmit 56 kb/s over a 100 kHz wide channel. $\endgroup$ Jul 23 '18 at 17:52

I know there's actually people working on a system for that band. It'll be SDR-based.

There's no technical restriction that says you can't have a couple hundred kilobits per second in a couple hundred kilohertz. It might just be the case that for commercial bands, the available bandwidth isn't attractive and regulatory restrictions make it unfeasible to implement a high-rate system. Since different regulations apply to Ham radio, this is solvable.

Glenn, however, raises an important point: In some regions, you simply cannot legally transport high data rates. That's kind of a stupid legislation, but it's how it is. In Europe, as far as I know, usage of this band is subject to bandwidth, but not data rate restrictions.


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.