Suppose they're doing a prop/system study? Also, why are they bothering the HAM bands with what is surely going to be commercial traffic in nature.
It has been mentioned here, and some of the folks over at QRZ forums are pretty emphatic, that this might be related to HFT experiments. I am skeptical of that. Over at QRZ and a few other forums they have gone as far as finding the experimental licenses that might be involved, but again I am skeptical.
Why the skepticism? No one really compelling reason, but lots of little things that push me away from any such connection.
No license found covers all of the frequencies observed. Between the 3 or 4 license being discussed they cover most, but not all, of the freqs seen, and no one license comes close to covering them all by itself.
The majority of the licenses seem to apply (based on emission types) to much wider bandwidth signals. I think only one of the licenses lists OOK or N0N modulation.
The most promising license was not valid until a month after I first logged these signals, and I have no idea how many other people saw them before I did.
So technicalities of the licenses aside, what about the signal?
These signals appear to be simple unmodulated pulses. There is no data that I can detect on the pulse itself.
It has been suggested that maybe the data is in the pulse timing, but the timing of these pulses is very consistent, they may, or may not, be timed at zero UTC second, but whatever their start time is they don't seem to vary much. All variations I have seen to date could be accounted for by propagation changes (altitude of the reflective layer, single hop vs double hop, etc)
It has been suggested that the data is in the frequency selection. However this would seem to me to be a very limited data set, and for a system that is supposed to be based on blindingly quick trades (advantages in milli, if not micro, seconds) having to get your next bit of data one full second later seems like a non-starter.
So really, I just don't see it being related to HFT unless it is very early in the process and they are still trying to figure out how to predict / quantify / resolve varying propagation delays. And if they were doing that, why such small frequency segments and steps? When they are active they are a couple hundred kHz wide hopping at most, and often sub 50 kHz across multiple freqs, what does that tell you that setting on one freq will not?
So, if not HFT, what is it? Hell, I don't know, ain't radio great?
What we really need to happen is a couple of other people take accurate time cuts on when the pulses arrive. We have two measurements, at least one more will give us enough data to do a TDOA calculation on the signal, and arrive at an approximate signal source location.
T!