We seek to understand and document all radio transmissions, legal and otherwise, as part of the radio listening hobby. We do not encourage any radio operations contrary to regulations. Always consult with the appropriate authorities if you have questions concerning what is permissible in your locale.

Author Topic: How does one Identify a Mystery Signal?  (Read 2745 times)

Offline K5KNT

  • Full Member
  • ***
  • Posts: 110
  • Ogden, UT USA
    • View Profile
How does one Identify a Mystery Signal?
« on: April 09, 2018, 1300 UTC »
What should one do to attempt to identify a mystery signal before posting to this forum? For sake of this discussion, let's assume the following:

  • The computer is running Windows 10.
  • He/she is a beginner in the hobby.
  • They only have access to on-line SDRs.
  • They are able to do screen recordings using OBS Studio.
  • They have Audacity and/or Adobe Audition CC audio editing software.
  • They have Adobe Premiere Pro CC video editing software.

I know that many of the on-line SDRs have labels to identify known signals, but how would you identify the signal if there is no label?

Is there any other software, free preferred, that is needed or would be very helpful? Assume that they don't want to spend a lot of money because they may not like the hobby.

What are the steps this person should take to identify this mystery signal?

Let's also assume that this signal will later be identified as a Link-11.

Would the process be the same of other types of signals?

Thanks,

Kent/K5KNT
K5KNT

Offline Strange Beacons

  • Sr. Member
  • ****
  • Posts: 264
  • Naples, Florida USA
    • View Profile
    • Strange Beacons Website
Re: How does one Identify a Mystery Signal?
« Reply #1 on: April 09, 2018, 1421 UTC »
One step that I would suggest is that mystery signals should first be checked out in the Signal Identification Guide. Yes, it can be time-consuming to do that, but more often than not, I end up discovering that the strange signal that I've heard/recorded can be found in that database.

Curt / W9SPY

Offline K5KNT

  • Full Member
  • ***
  • Posts: 110
  • Ogden, UT USA
    • View Profile
Re: How does one Identify a Mystery Signal?
« Reply #2 on: April 09, 2018, 1446 UTC »
Ok. That and the HFU Wiki are both good resources. What If I want to learn how to analyze/identify a signal myself? What steps should I take? I could then compare my findings with those resources.

Edit to add: I took information from a youtube video of a link-11 signal on 6243kHz and entered the following in the SIGWiki search database form:  Frequency range: "6240-6250", Mode: "USB". Using those search parameters did not bring up any results.
« Last Edit: April 09, 2018, 1454 UTC by K5KNT »
K5KNT

Offline Token

  • Global Moderator
  • DX Legend
  • *****
  • Posts: 2118
    • View Profile
Re: How does one Identify a Mystery Signal?
« Reply #3 on: April 09, 2018, 1525 UTC »
Don't always believe the SDR labels for signals you are not familiar with, I see a lot of wrong labels or labels for transient signals on SDRs that might mislead a listener.  So if the label says it is XYZ signal use that as a starting point to research what it might be, but don't assume just because a label on an online SDR identifies the transmission type that it is correct.

First step is to note time, date, frequency, receiver mode, and receiver location.

An audio recording should be made, exactly how to make the recording, what receiver conditions to use, depends, a little, on the signal.  But best default mode for unknown signals is USB and a filter setting wide enough to capture all of the signal.  Name the file in a meaningful way.    An example file name might be something like " Unk_1PPS_10149kHz_USB_17022018_1628z_local.mp3 "  This is my naming convention for files at home, it tells me what the signal is (Unk_1PPS), the RX frequency (10149kHz), the RX mode (USB), the date (17022018, DDMMYYY), the time (1628z, 1628 UTC time), and that it is a locally received signal.  If it had been received on a remote the "local" at the end would be "rem_xxxxxx" with the x's being the location or name of the remote.  Or if the file has no location on the end I assume it is local.

If someone may do analysis of the recording save it in the best format available, one that compresses the least.  WAV 16 bit PCM is one I use for those.  However, the files can get pretty big, so for files I am posting to the web as examples I use MP3 or some other format with moderate compression to reduce file size.  It might be worth it to save both a WAV and a smaller MP3 file, that way you can post the MP3 and have the WAV locally if more is needed.  That is what I do, I save one WAV and one MP3 of each signal, one to keep (the WAV) and one to post (the MP3).  Regardless, if you save as a compressed file don't over compress and start introducing bad artifacts.  Play around a little with your specific software and find a compression level you like that does not seem to mess with the audio much.

Assuming using an SDR, if possible grab a screen shot of the signal on the waterfall.  Get as much detail as possible.  Don't forget to capture things like the frequency scale on the SDR so that people looking at the screen shot have a reference to the signal bandwidth and what is going on around it.

At the most basic level with Win 10 you can hit Alt-Print Screen and paste into MS Paint, that will grab a screen shot.  If you have other software that allows more versatility so much the better.

Videos are great, and all the same stuff above applies.  If you don't know what mode the RX should be in start with USB, make sure the RX filter settings capture all of the signal, make sure the audio is not overly compressed, make sure the video shows some detail such as frequency scale of the SDR.

All that gets the signal in an archived way for you to either analyze it later or share it for others to identify.

Beyond that you can start trying to ID features yourself, but how to do that is going to vary depending on the software you have.

I don't know what software works under Win 10, I don't use a 10 machine for any analysis or recording.

A program like Spectrogram 16 will allow you to see the signal in greater detail, zoom in on features and measure frequency components.

A program like Signal Analyzer will allow you to take the signal apart in greater detail, but it is not free and I am not sure of the current status of the software.

A program like Soundcard Scope will allow oscilloscope like viewing of signals.


Sorcerer is an orphaned software with some ability to demod common modes.  The legal question on this software is in question, it is not and never was freeware, but I don't think the company exists anymore.

Along with various web pages (https://www.hfunderground.com/wiki/Signal_Identification , https://www.sigidwiki.com/wiki/Signal_Identification_Guide , http://sferix.myweb.hinet.net/hfasia/ , http://signals.taunus.de/ and many more) there are some things in print that can help you also.  Technical Handbook for Radio Monitoring HF is one, while a bit expensive it has many examples of modes and images and tables to help ID signals.  Another is Signal Analysis for Radio Monitoring, unfortunately it leans heavily on expensive software but as a guide you can apply the information to other software sets.

T!
T!
Mojave Desert, California USA

Offline Token

  • Global Moderator
  • DX Legend
  • *****
  • Posts: 2118
    • View Profile
Re: How does one Identify a Mystery Signal?
« Reply #4 on: April 09, 2018, 1534 UTC »
Ok. That and the HFU Wiki are both good resources. What If I want to learn how to analyze/identify a signal myself? What steps should I take? I could then compare my findings with those resources.

Edit to add: I took information from a youtube video of a link-11 signal on 6243kHz and entered the following in the SIGWiki search database form:  Frequency range: "6240-6250", Mode: "USB". Using those search parameters did not bring up any results.

There is no web site that does it for you.

Using your Link-11 example, Link-11 is a military mode.  Further, it is a tactical military mode, meaning it adapts and changes as needed.  OK, the mode itself does not adapt and change, but the platforms using it do.  This means that while there are some well known and long standing Link-11 freqs there are many more Link freqs that show up for a few days/weeks and then go away, never to be seen on those freqs again.

So you often cannot search for it by freq.  You have to learn its features and recognize it when you find it.

You need software and an ear to distinguish features, you focus on the most unique and identifiable features, and then you match that to ID.  For example, Link-11 CLEW has a Doppler tone at 605 Hz and a Sync at 2915 Hz, along with 14 data channels or tones.  Both the Doppler tone and Sync are very indicative.  And this is what gives CLEW its easily heard dot dot dot dot daaasssh sound.  Histogram software can show you the spectrum of the signal, and allow you to ID based on key features.

In the case of something like CIS-12/20 (Russian AT3104D modem) there is a Doppler tone at 3300 Hz and 12 (or 20) data channels, all easily seen on a histogram.

T!
« Last Edit: April 09, 2018, 1537 UTC by Token »
T!
Mojave Desert, California USA

Offline K5KNT

  • Full Member
  • ***
  • Posts: 110
  • Ogden, UT USA
    • View Profile
Re: How does one Identify a Mystery Signal?
« Reply #5 on: April 09, 2018, 1610 UTC »
Original reply deleted as I had not seen Token's first reply.

I have both books Token mentioned, but I'm not sure where to begin. I'll look into other software and post in the proper forum about any that I find.

Maybe Link-11 wasn't the best example, but I "think" I'm understanding what you are saying.

Kent/K5KNT
« Last Edit: April 09, 2018, 1621 UTC by K5KNT »
K5KNT

Offline Josh

  • DXing Phenomena
  • *******
  • Posts: 4322
    • View Profile
Re: How does one Identify a Mystery Signal?
« Reply #6 on: April 09, 2018, 1638 UTC »
Once you're able to discern between fsk, mfsk, and psk signals by ear you can then more easily use tools like Token mentioned above. Here's a link to a pdf on how to use SA;
https://dk8ok.files.wordpress.com/2015/09/sa_steps.pdf

As hobbyists we have to do a lot of signal id work by eye/hand/ear, the pros have software (and limitless government funding) that can do it for them.

Behold;
https://www.youtube.com/watch?v=wgHMUSE-Z-U
https://www.youtube.com/watch?v=H8UXhyHgpm0
We do not encourage any radio operations contrary to regulations.

Offline Token

  • Global Moderator
  • DX Legend
  • *****
  • Posts: 2118
    • View Profile
Re: How does one Identify a Mystery Signal?
« Reply #7 on: April 09, 2018, 1701 UTC »
It is tempting to try and research a signal by frequency, sometimes this is very valid, other times not so much.  The problem, of course, is what to use as a reference?

A quick and easy, searchable, ref for possible signals is the NSA bot in the #wunclub IRC.  You can search by freq or you can search the text.  Of course, this information is based on user inputs to the NSA bot, and in recent years user inputs have been very lacking.

The UDXF web site (  http://www.udxf.nl  ) also contains past logings from multiple sources, while an excellent resource it is not in a neat and tidy data base to search.  You must download the various PDF files and then do something with them, search them, import them into another program, etc.  Again, all this information is based on user inputs (to the UDXF group).  So while it is very good, accept the fact that it might be someones best guess for a specific entry.

And the Global Frequency Database at Globaltuners is a decent, searchable, resource ( http://qrg.globaltuners.com/ ), but again, not complete or 100% accurate.

T!
« Last Edit: April 09, 2018, 1703 UTC by Token »
T!
Mojave Desert, California USA