31 August 2015

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.

28 August 2015

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.

27 August 2015

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 


26 August 2015

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.



25 August 2015

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

23 August 2015

R.Romania Int. - Tiganesti: a DRM-B waveform


This is a DRM B-version: OFDM 206-tone (+ 1 empty place) 46.8Hz channel step, baudrate 37.5Bd and k=1/4 as reported in the radioscanner.ru page for DRM-A (bottom of the page, values of the k parameter). The official DRM System reference publication is ETSI DRM System specifications

OFDM symbol parameters as measured

A DRM trasmsission super frame (TSF) consists of three channels, or three transmission frames:
  •  FAC, Fast Access Channel: The Fast Access Channel is used by the receiver to obtain information about the OFDM signal properties and the SDC/MSC configuration
  • SDC, Service Description Channel: The SDC contains the information needed for MSC decoding, like the multiplex frame structure, as well as other additional information
  • MSC, Main Service Channel: audio and data stream
Since one DRM frame is 400 mS, the duration of a TSF is always 1.2 seconds. 



According to the ETSI DRM definitions, the OFDM carriers are divided in two groups and indicated as follow:


Some tones within the OFDM transmission frame are modulated with known fixed phases and amplitudes. These care pilot tones for channel estimation and synchronization. The positions, amplitudes and phases of these tonse are carefully chosen to optimize the performance, especially the initial synchronization duration and reliability. These tones are used by the receiver to detect the presence of the received signal and to estimate its frequency offset. They may also be used for channel estimation and various tracking processes. 
In DRM-B there are three frequency references, which are 750 Hz, 2250 Hz and 3000 Hz as referenced to the DC carrier that is not used; the tone numebers are: 16, 48 and 64 according to ETSI numbering (tones 120, 152 and 168 in SA numbering).
Looking at these pilot tones, they seem to have multiple modulation outlines that depend on the number of sent symbols:

- PSK-2 (periodicity 2)
- PSK-4 (periodicity 3)
- unmodulated (periodicity 4)
 
pilot tone 120 (ETSI 16)
pilot tone 152 (ETSI 48)
pilot tone 168 (ETSI 64)
The two couples of tones indicated by the ETSI numbering as -103,101 and -101,103 (1,205 and 3,207 in SA numbering) belong to the set of "gain reference tones" (those tones have a power gain of 2) and those who are close to the band lower and upper edges are over-boosted by a further power gain of 2, as it is clearly visible below. The periodicity of the gain reference pattern is 3 symbols:


Modulation is QAM-64 in even symbols:


in odd symbols modulation is QAM-64 with the evidence of outer phases (service and control) that are sent each three symbols; note that in this picture the signal has been resampled at a different frequency:


The positions of the service and control 'cells' are shown in Annex-A "Pilot reference" of the ETSI DRM document and justify the behaviors seen above.



Juts two words about the software "DREAM".
DREAM is an Open-Source Software Implementation of a DRM (Digital Radio Mondiale) Receiver under the GNU General Public License (GPL), it works both on-line and in post-production by processing wave recordings. Below some screenshots of DREAM that is struggling with same file processed by SA.
The software provides many measurements in the Chart Selector menu, including the views of the three phase constellations of DRM channels.


waterfall
Power Spectral Density
MSC-SDC-FAC constellations

DREAM links:
http://drm.sourceforge.net/wiki/index.php/Main_Page
http://sourceforge.net/projects/drm/?source=typ_redirect





15 August 2015

Rohde & Schwarz ALIS (RS-ARQ selcall)



This is the Rohde & Schwarz proprietary standard "ALIS" (Automatic Link Setup) selective call, an adaptive ALE procedure based on FSK modem and ALIS processor. 
Quoting Hoka: "Rohde & Schwarz simplex ARQ, so far found in use by German, Italian, Nairobi and Turkish Diplo services, typically 228.7bd but reports of 457.0 have been noted. There appears to be no "real" name for the data system. Some people call RS-ARQ as ALIS but strictly speaking, ALIS is only the automatic link processor and frequency management system. It is not responsible for generating the traffic. ALIS is therefore somewhat of a misnomer. The modems generating the traffic are the GM857 and GM2000. Our suggestion is to stick with RS-ARQ as the system name."

The signal has been heard today afternoon at 1620z on 12270.5 KHz on USB. The waveform is an FSK 228.6Bd/170 (normal FSK modulation, ALIS basic feature), ACF value (in the analyzed sample) of 342.6 msec or 78 bit. By using the SA "scan raster" method is visible the feature called "diagonal bit" that can be used to recognize the signal.


Fig. 1

The following basic parameters are permanently stored in the communication processor unit:

- address list with geographical coordinates and distances of corresponding stations;
- frequency pool (total number of frequencies up to 100, with a maximum of
16 frequencies for each pool and a maximum of 25 different pools);
- date and time;
- sunspot number.
The calling (master) station establishes the connection. It transmits a defined number of frames on each frequency of the desired pool to enable a weighted bit addition at the receiving station. The frames are repeated because the slave station must have enough time to scan through all the programmed frequencies and receive at least three frames on each single frequency for synchronization.
After the correct reception of (at least) three frames from the master station, the slave station transmits a synchronization acknowledgement and additional information about the reception quality along with the monitored status of the sending station to ensure that the two stations are working in the same operation mode.
 

The selcall phase maybe followed by the data transfer phase which is performed using several waveforms, as indicated in in the field "followon"  ("status") of some decoded selcalls (Fig. 3). Note also the different "pool size" which is related to the number of the allowed frequencies for the current pool.


Fig.3

14 August 2015

CIS-60 idle


The signal, heard on August 12.08 16810.0 KHz on USB, is a variant of the CIS-60 idle, possibly for test or adjustment/calibration purposes. CIS-60 waveform is analyzed here http://signals.radioscanner.ru/base/signal231/  and here CIS-60 in this blog.
I say a "variant" since a similar test signal is also mentioned in radioscanner.ru at that same address, but in my sample, at a first glance, looking at its sonogram the 60 tones seem to be housed in what looks like a curious 40+20 (+1 pilot tone) grid: although all the tones are not modulated, the top 20 tones (41-60) appear in some way shifted with respect to the lower ones (1-40). 
After resampled the signal at 8888.88 Hz (shift * 200), from 1-40 tone to tone there is a 180º phase shift in the relative constellation, but,from tone 40 to tone 41 there is no variation (Pic. 1). As AngazU too suggests, this could point to a different phase of the top 20 tones during the initial generation of the signal. Moreover, looking at the OFDM parameters (Pic. 2), there is a different baudrate in (at least this) test mode (30Bd) with respect to the usual traffic mode (35.5Bd). BTW, a 30Bd traffic mode was also noted.
Recordings available on request.

Pic.1 - phase variation

Pic. 2 - OFDM parameters

10 August 2015

Unid MFSK-11 waveforms

wf1
wf2

Heard today on 12167.0 KHz USB starting from 1250z, the recordings of these two signals are available on request. The upper waveform (Pic. wf1) exhibits 11 tones, step between tones 250 Hz and speed = 125 Baud x carrier quite the same as MIL 188-141 ALE. 
The second waveform in Pic. wf2, more complex, has segments in which are sent only a certain subset of tones.



It's difficult to say anything for sure about it, this signals is reported as CIS MFSK-11 here and classified as "unid" in an "old" post (October 2013) by Karapuz in radioscanner.ru. Browsing the web, only some reports of receptions and nothing else about identification and users.
Below a complete dump of a decoding of a short tramission, the first that has been heard:

  

SERDOLIK Selcall


This waveform provides a 31 frequency grid, 40 Hz spaced, and 40 Baud rate (pic. 1). The signals has MFSK insertions into the sync-marker sequence
The call shown in pic. 2 ends into a sync-marker sequence and since no data follows, this could be a sort of null call. 
In pic.3 a complete session, 105 secs lasting, with the call followed by data sent in MFSK-34 mode (CROWD-36): difficult to state if this 'traffic' corresponds to user-data or to link-setup informations with the actual traffic sent in another frequency and/or mode.
The recording in pic. 4 is a more complete example: the call and the sync timing, the sending of link-setup information through MFSK-34 and then the transmission of data through the OFDM 35-tone BSPK waveform. It's interesting to note that the both the call and the MFSK-34 two initial tones exhibit a shift that is the half of the usual.

Pic. 1
Pic. 2
Pic. 3

Pic. 4