updated
Unid transmissions heard on 8120, 8130, 8140, and 8150 KHz (the latter moved to 8160 Khz) thanks to the KiwiSDR located at N4BUT Orlando, FL [1].
 |
| Fig. 1 |
At first sight the signal seemed a MPSK modulation, but working the signal along with my friend cryptomaster some other interesting features came out. Each transmission consists of four 100 Hz separated channels (a,b,c,d) each occupying a band of about 900 Hz, for a total bandwidth occupation of about 3900 Hz (Figure 2).
 |
| Fig. 2 |
In turn, each of the 4 channels consists of 4 sub-channels with a modulation rate of 49.6 Baud, the used modulation seems to be ASK2/OOK [2].
 |
| Fig. 3 |
Each of the four sub-channels shows strong ACF peaks of 1290 ms corresponding to a 64-bit length frame: the "aggregate" frame therefore has a length equal to exactly 1 Kb, ie 1024 bits or 128 bytes (64x4x4).
 |
| Fig. 4 |
Below in Figure 5 is a comparison of the four channels obtained from the analysis of my friend cryptomaster.
 |
| Fig. 5 |
Although I have kept an eye on that initial portion of the 8 MHz band (fixed/mobile band, shared with marine for simplex purposes), those transmissions have not appeared again (at least until today): difficult to define their purpose and user(s).
Below the interesting comment sent me by my friend Nicola, who I thanks for the collaboration:
"The interesting signal discussed in the blogpost “Unid ASK2/OOK
transmissions” dated 1 February is probably a Frequency-Time Matrix
(FTM) system, where a data symbol (bit string) is represented by
FT-matrices, i.e. combinations of frequency and time domain
positions. The 'secret' of this robust mode is that no frequency is
repeated within each matrix. This characteristic is used to enhance
the resilience against frequency domain broadband noise (or fading)
affecting a broad range of frequencies and time domain narrowband
noise (or fading)."
https://disk.yandex.com/d/7u74pzY-bPntXg
[1] http://sdr.n4but.com:8173/?f=8120.00iqz10&pbw=10000
[2] https://en.wikipedia.org/wiki/Amplitude-shift_keying
This is a signal I call the MFSK Oddity in my logs (it was first described to me as an MFSK signal). I have found it on dozens of frequencies from below 7000 kHz to above 22000 kHz. I have found it with several different speeds / rates, as well as at least 5 different bandwidths, from 2.6 kHz to over 10 kHz. One source claims it is related to HFT. I have literally hundreds of recordings of this signal taken over the last few years, and the format has slowly changed over time. I suspect the data is in the frequency domain, not the amplitude domain.
ReplyDelete