It wasn't my intent to have this next post be about locating a source of a transmitter - but the temptation proved irresistible.
Over the past several days I'd been working on an addition to the Northern Utah WebSDR: A temperature-based frequency control of the local oscillators on some of the receive chains. The receivers in question are based on the Si570 synthesizer and are prone to temperature-based frequency drift, and since they have internal reference oscillators, there is no way to externally lock them.
For this temperature-based stabilization to work, I have correlated the room temperature with the actual frequency, so I have been frequenting the bands/receivers with the aforementioned issues and making measurements - but I digress: It was during this activity that I noticed this massive signal at the top of the 20 meter band, occasionally firing up and clobbering ongoing conversations by U.S. amateurs.
Figure 1: Waterfall display of the signal around 14.350 MHz. Invisible are ongoing QSOs underneath this strong signal. Click on the image for a larger version. |
What is it?
Upon seeing this, I had my suspicions - but I fired up the TDOA (Time Direction of Arrival) system on the KiwiSDR network, using five receivers within the zone of reception scattered across the continental U.S. Multiple sessions of direction-finding over several days yielded similar results to this map:
This clinched it - it was likely shortwave-based high-frequency trading.
Who are they?
As you may (or may not) know, the so-called "High-Frequency" trading utilizes the very small differences in the prices of trading instruments (stocks, etc.) that occur over time. The idea has nothing to do with "HF" like shortwave radio, but rather it is the notion that if one can buy or sell a tiny fraction of a second before someone else, differences in prices may be exploited. One of the aspects of this type of trading is that conventional means of data transport (e.g. fiber optics) is too "slow": Light travels at about 1/3 the speed as in open air through a glass fiber and this means that compared to a radio wave on a "direct" path, data transmitted via fiber will arrive later - and this does not include delays due to the equipment in that data network.
What this means is that some entities are experimenting with the use of the HF bands for the most direct, point-to-point means of conveying this information possible - and it seems that some of this information is being transmitted on amateur bands, as the above indicates.
Not surprisingly, these entities are very secretive - but others have done a bit of digging in public, FCC databases.
Here are a few links:
- ARRL News article about "High Frequency Trading".
- QRZ thread about "High Frequency" trading.
- "Sniper in Mahwah and Friends" article about High Frequency Trading.
As noted in the QRZ thread, the Part 5 experimental license frequency includes the entirety of the 20 meter band, with no requirement for identification.
While many amateurs seem to be surprised about this, I was not: There are several instances where Part 5 licenses have been issued (I can provide an example via email) - the applicant providing frequency ranges in their application that encroach on any number of other services - and been issued permission to operate there - but there's typically a caveat: They are not to interfere with existing, licensed services.
It's this last point that's a bit tricky. Anyone that has operated on HF knows that this is a dicey proposition as it's entirely possible that other users of a particular frequency may not be able to hear - or be heard by - the "offending" station. As an example, if station "A" and "B" are in QSO - but the offending station can only hear - or be heard by - station "A", it cannot "know" to avoid transmitting while station "B" is transmitting. It would seem that those who make the rules have overlooked this particular of aspect of HF propagation when it comes to utilizing HF "whitespaces"(e.g. seemingly-unused frequencies.)
"I've been getting QRM'ed - what can I do?"
The complete list frequencies on which these operations are currently unknown - and the fact that they are not assigned specific channels may make such information impossible to know other than by direct observation. So far, the two frequencies of which I'm aware is that depicted above (around 14.350 MHz) and another around 4.4 MHz - but I have little doubt that there are others: If you spot similar signals on other frequencies, please comment.
If you note similar interference issues, please contact your amateur radio representative. In the U.S., you may contact the Volunteer Monitor program at the ARRL (see information here.) Unfortunately, a quick search did not reveal any specific contact information regarding this program: If you have such information, please let me know via a comment.
* * *
Update:
"Luke" noticed this post and tweeted it, emailing me a few links: Here's a bit of information others have dug up:
- Delmarva Broadband is authorized 14.35-14.99 MHz at 186 kW. Contact is a law office in DC. See apps.fcc.gov/els/GetAtt.htm
- … and M-Wave is authorized the 14-14.99 MHz at 16 kW. See apps.fcc.gov/els/GetAtt.htm