SERDOLIK OFDM 35-tone BPSK

Pic. 1 - Serdolik OFDM 35-tone

here a signal that is another variant of the Serdolik OFDM 35-tone: I had a similar waveform (40 Baud and 50 Hz channel separation) with PSK-4 modulation, this waveform shows the same values of speed and channel separation but the modulation in absolute/relative positions is BPSK (Pic. 2). Signal has been resampled at 8000 Hz. The same waveform is also reported in this discussion in radioscanner.ru forum. 

Pic.2 - OFDM parameters and modulation in channels

Moreover, phase in carriers #1-2-3 and #33-34-35 does not change in relative constellation then they are not modulated altough they could carry service informations. Looking deeper at the signal, in this case I used the software tools "baudline - signal analyzer" (http://www.baudline.com/), some channels in time change their modulation (Pic. 3) and it would take a much more in-depth analysis.

Pic. 3 - the same signal as seen in baudline

The signal was heard today (31 August) in the morning at 0750z and tuned on 16120.8 KHz on USB: the recording is available on request.

CIS-20 in idle state

The main features of the waveform, as reported in radioscanner.ru, are: k = 3/5, 20 channels + 1 pilot, channel modulation pi/4 DQPSK, speed 75 Baud and channel step 120 Hz.

In this sample the modem was heard when in idle state: modulation switched to BPSK (4 positions in the absolute constellation) but same manipulation speed and channel separation (75 Baud and 120 Hz). Notice the resampling at its native frequency (7200 Hz as shown here). ACF in idle state is 200 mSec.

 

AT-3104D waveform (idle mode) OFDM parameters

AT-3104D waveform (idle mode) ACF

The recordered signal does not have a good quality and modulation is a bit confused, so I isolated a single channel to confirm the relative constellation (BPSK).

AT-3104D waveform (idle mode) single channel inspection

 As other characteristic, at least in this sample, is the carrier #13 that is not modulated

According to this post in radioscanner CIS-20 is a new version of CIS-12 so probably sourced from the same AT-3104 modem (or the older AT-3004D). Both the two modems were designed to have many different devices plugged into them to form the modem’s input – most often teleprinters unique to the different military services and civilian agencies that use them, and also, vocoders and speech encipherment systems, but all are transported in the same fashion – using one of these 2 waveforms.

hurry up and get moving!

how to use an AT3004-D (CIS-12) modem as "beep-beep-beep" to draw the attention of the (lazy) operator at the other side of the link...
https://www.dropbox.com/s/rj6cw0yqkefsqf6/26-Aug-2015-233242%206.211%20MHz.wav?dl=0 

CIS-45 v2 HDR modem 40 Bd π/4 DQPSK

This is another variant of the CIS-45 v2 waveform exhibiting the π/4 DQPSK manipulation in the channels and 40 symbols baudrate: this way the operating speed is increased to 3600 bps.

prob. V22 Chinese number station

(updated)
As reported in http://www.numbersoddities.nl/Chinese-stations.pdf "it seems that V22 station have been replaced by a digital mode(s). A so far unid mode very similar (if not equal) to PSK-63F (BPSK 62.5 Bd) used by HAMs has been noted on V22's frequencies but it cannot be decoded by HAM software neither in BPSK-63 nor in PSK-63F". 

There are no evidence or proof about the source/user which are behind these waveforms, only "rumors" as those reported in the above document. Anyway, me and Karapuz heard what is possibly the V22 station, or a "Chinese number station", running in BPSK at 62.5Bd and in DQPSK at 62.5 and 250 Bd on 9611.5, 14941.0 and 15799.0 KHz USB/cf, as sumarized in the pictures below.
All the recordings, as usual, are available on request for your analysis.

BPSK 62.5 Bd, 14941.0 KHz

DQPSK 62.5 Bd, 15799.0 KHz

DQPSK 250.0 Bd, 9609.5 KHz