Thanks to SDRplay, I was sent both their new RSPdx and older RSPduo SDRs at the end of January.
The main reason was to get them integrated into Procitec’s go2MONITOR and go2DECODE software, to increase the number of SDRs that the company’s products are compatible with.
This I’ve been successful in doing with the RSPdx – I’m still to unbox the RSPduo at this time of writing.
First of all though, I’ve been extremely pleased with the RSPdx in its own right. The SDRuno software works really well, is pretty easy to use – and it looks good too.
The fact that you can have up to 10 MHz of bandwidth is brilliant, and it isn’t too bad on the CPU usage either – running at around 25% with 10 MHz bandwidth on my ancient PC. Used with SDRConsole you can cover a good number of frequencies at once, and can record them if necessary. Of course, you can do this with SDRuno too, but at the moment only IQ – you can’t record individual frequencies.
Saying that, I’ve seen the SDRuno Roadmap for future releases and not only will recording of individual frequencies be possible, a more advanced scheduler is to be included. This is something I feel SDRConsole – amazing though it is – is lacking when it comes to single frequency recording. There is also the issue with SDRConsole that you are limited to recording only 6 hours worth of wav file per frequency.
Anyway, I digress. Back to the RSPdx and go2MONITOR.
To get the SDRs to work correctly with any of the go2 products means creating a configuration file and adding a ExtIO DLL file to the software. This is reasonably easy to do once you get use to it and it enables a GUI to become active so that you can control the SDR through go2MONITOR.
One interesting aspect with the RSPdx GUI is that regardless of what you enter as some of the parameters in the configuration file, the ExtIO file overrides these. Effectively, I just left some of the data as found in a basic configuration template and let the GUI do all the work for me.
So, below are some of the results with today’s first test.
First of all I went into the VHF/UHF side of things and targeted the local TETRA networks. These were found easily and after messing around with the GUI, I was able to get go2MONITOR set up to nicely find all the emissions within the 1.6 MHz bandwidth I’d chosen to use
From there, all I had to do was to select one of the found emissions and let the software do its thing.
Next I moved on to HF where there’s a plethora of data to choose from to test out the SDR. There was quite a large storm going through at the time and my Wellbrook loop and coax feed were getting a bit of a bashing with some considerable interference being produced with the really strong gusts, as can be seen below – the interference between the two HFDL bursts is one such gust.
I’ve frequently mentioned the Results Viewer that’s part of go2MONITOR and with things such as HFDL and TETRA, that process data quickly from lots of signals, this part of the software comes into its own.
The image below is two minutes of HFDL monitoring. All the red blocks is received data that scrolled through the Channel window too quickly to read live. In the viewer you can select any of the signals and you’ll be shown the message as sent. In this case, it is one sent by an Open Skies Treaty observation flight OSY11F.
By looking at the Lat/Long and comparing it to the flight history from FlightAware and its location at 1313z it ties in nicely. This flight was carried out by the German Air Force A319 1503 specially kitted out to make these flights.
go2MONITOR has a basic map function within the Result Viewer function so if there’s any Lat/Long position within any message it will plot it – as shown below for OSY11F at 1313z.
Within the General tab of Result Viewer you can get all the parameters of the signal.
One final test that I carried out was how well everything coped with a bigger bandwidth. In HF I can use up to 3 MHz of bandwidth with the licence I have – going up to 10 MHz once into VHF/UHF. In HF then, I selected 3 MHz in the GUI and then ran an emissions search.
My PC is nearing the end of its life but it coped easily with the amount of data found despite only having 4 GB of RAM with a 3.6 GHz AMD processor – a new PC is in the pipeline that is going to give me much better processing power.
Despite having 3 MHz available, not everything was identified. Most of this was at the fringes of the bandwidth, but some of the weaker signals also failed. That doesn’t mean you can’t then process them further, you can, it’s just the Emissions scan hasn’t quite been able to ID them. Saying that, the software managed to ID things within 2.2 MHz of the 3 MHz bandwidth.
I picked one of the weaker signals to see how both the RSPdx and software coped and they did very well, pretty much decoding all of the CIS-50-50 messages that were coming through on 8678 kHz.
So, overall, pretty pleased with how the RSPdx works with go2MONITOR.
Once I get a better PC I’ll be able to test at bigger bandwidths but even with 3 MHz here I was able to achieve the same, if not better, results than I have with the considerably more expensive WinRadio G31 Excalibur I have been using previously (running with the G33 hack software).
Not that I’m likely to really use go2MONITOR at big bandwidths – 1.6 MHz is probably fine for me – but for Pro’s there’s no doubt that having these “cheaper” SDRs would make absolutely no difference over using an expensive one such as those in the WinRadio range. In all honesty, I don’t think I’ll be holding on to the WinRadio for much longer – I’m more likely to get another RSPdx to cover this area of my monitoring.
On its own, as an SDR, the RSPdx is worth the money I’d say. I like it just as much as I do the AirSpy HF+ Discovery – the only real difference I can see between these two SDRs is the max bandwidth available.
I recently completed an article for Jane’s Intelligence Review magazine on the activation in December 2019 of the Russian Over-the-Horizon Radar system (OTHR) 29B6 Konteyner near Kovylkino in Mordovia.
Like all of the articles I write for them, many parts and imagery are removed due to space constraints in the magazine – for example, see my previous blog on the Murmansk-BN EW system where I have been able to add a substantial amount of extras that couldn’t be published. So, whilst I can’t publish here the actual article on Konteyner, I can show some of the extras that were removed.
How OTHR works
I could go into how OTHR works, but it’s been covered elsewhere in extreme detail. One of the best places for a basic overview is Wikipedia, where the image below is taken from.
Officially designated Object 5452, construction work of the original transmitter and receiver sites commenced in 2000, taking two years to complete.
The Konteyner receiver site, with one array, was situated 6 km to the South West of Kovylkino, whilst the transmitter site – also with one array – was located 5 km north of Gorodets in Nizhny Novgorod Oblast. The system covered airspace to the west of Russia with a central bearing of 275 degrees, arcing out in a fan, with an approximate range of 3000 km (depending on radar pulse rates – covered later). Due to Ionospheric bounce a null area is created that is approximately 900 km in depth from the transmitter site. Here, nothing would be picked up by the Konteyner systems, and other OTHRs such as Resonans-N and standard Air Defence radar systems are used to fill in these gaps.
However, the Gorodets site is no longer in use, despite many blogs and expert publications saying otherwise – Jane’s included (until my article). Located at 56°41’34.1″N 43°29’11.3″E, this site has been dismantled since at least 6/2/18 according to Google earth imagery. All the concrete footings remain, but the antenna array is no longer there.
The receiver site at Kovylkino is still there, and from June 2016 construction had begun on two other receiver arrays, creating a triangle. Array one continued to cover a 275-degree bearing whilst the new arrays covered 155 degrees and 215 degrees.
Each receiving array contains 144 masts, all approximately 34 metres in height. They are split into three sections where the two outer ones – consisting of one group of 23 masts and the other of 24 – is between 300 and 310 metres wide. Each antenna here has 14 metres of spacing between them. The inner section contains the remaining 97 masts with 7 metres between each. The total length of the array is over 1.3 km.
Matching these receiver arrays was a new transmitter site just 15km to the South East. Imagery on Google Earth from 29/6/16 shows that there are three arrays being constructed in a Y pattern, each with the same three bearings as the receiver site. By 18/8/17 it is clear that the southern array originally thought to be covering 275 degrees instead covers 095 degrees. A second array is visible being built back to back with the 095 array to cover 275 degrees. Moreover, this meant that the original 275 degree receiver array was also being used by both transmitters.
The closeness of the transmitter site to the receiver site for long range OTHR systems is a strange one. In general they are a good 100 kilometers apart – the Australian JORN system is good example of this. Moreover, putting all the arrays so close to each other – at both sites – opens the whole system up to being destroyed, or put out of action, by just one air strike!
Each transmitter array has up to 11 generator buildings located to the rear of the antennas. Four of these buildings are also located at the original 095/275 degree receiver array. Google Earth imagery from 24/2/18 shows both sites still under construction. From 1st December 2018, combat testing of Konteyner had started and satellite imagery shows all four arrays to have generators in place by November 2018.
The transmitter site consists of 44 masts in a line, 500 metres in length. The masts themselves are of differing height with the 22 tallest ones approximately 34 metres tall. The remaining 22 are approximately 25 metres in height. The masts are split up into groups of 11 of each kind.
With ranges of over 3000 km for each transmitter – effectively there are four OTHRs in use – the number of radar tracks that are captured will be in their thousands, many of which are civilian. Moreover, static features such as large buildings are also captured, showing as background noise or unknown tracks.
There are two methods used to eliminate the background noise. Firstly, during testing many of these will show through time and are deemed static and can be filtered out. Secondly, this type of OTHR – known as OTH-B or Over-the-Horizon Radar (Backscatter) – employ a Doppler effect to distinguish between static and moving targets requiring fast computers with high processing power. Doppler uses frequency shift created by moving objects to measure their velocity and so can track targets travelling at any speed, even down to 1 or 2 knots for ship traffic. Whilst older Russian OTHRs – and likely Konteyner in its early days – would have struggled in this area, modern computers can cope with the Doppler methodologies used. Anything deemed not moving by the Doppler effect can be eliminated as a potential threat or track, and are also filtered out.
To further eliminate any overloading caused by unwanted tracks, areas of interest are set up within the radar coverage which are then further split into smaller areas or “search boxes” where radar returns outside of these are ignored. These search boxes can be moved by operators as required.
The radar system is unable to determine any height parameters therefore each track is just a target at an approximate GPS position, and could be on the ground or anywhere up to 100 km in altitude! In other words, it is the equivalent to a primary track in the standard radar world. Moreover, each track could be displayed at an operators console with a radar return that depicts the target to be kilometres in size! This further complicates determining the actual location of the track.
Finally, OTHR technology does have another drawback that is much harder to filter out. Just by looking at the images below you can see that a substantial number of aircraft tracks are still captured within the search boxes, particularly in busy airspace such as around airports and heavily used civil ATC airway systems.
One thing that OTHR doesn’t have is an Identification Friend or Foe (IFF) capability. Without IFF, this then makes it even harder to determine who is friendly, who is just an airliner or who is a potential threat.
Each of these tracks needs to individually interrogated and the routes plotted to eliminate the potential threat. For instance, all traffic into Istanbul pictured above tends to fly the same routes in and out of the airport there, so whilst the track can’t be fully removed from the display (or filtered out) it can be “ignored”. If IFF was an OTHR capability – and this is the same for other OTHR systems, not just Konteyner – then known transponder codes allocated to airports/airway systems etc. could then be filtered out. This happens in everyday ATC operations where certain transponder codes can be filtered out to remove clutter at the press of a button.
This then can make OTHR monitoring reasonably labour intensive for operators covering areas of high aviation activity despite modern computer technology being there to help.
OTHR range capabilities are controlled by the pulse rate of the signal sent by the transmitter site. In general, Konteyner operates at 50 pulses per second (pps) giving a range of approximately 3000 km. This pulse rate is also used by many other OTHRs such as the Australian JORN system (Jindalee Operational Radar Network).
OTHR has a potential advantage over standard radar systems in that it can track stealth aircraft such as USAF B-2s and F-35s. JORN reportedly tracked a USAF F-117 Stealth in the 1990’s that was on a round the world flight proving it couldn’t be picked up by radar! The Royal Australian Air Force (RAAF) were so confident they’d tracked it, they gave the details of positions the F-117 took to the USAF. I couldn’t find any confirmation on this from USAF documentation but it is possible.
By using the Ionospheric HF bounce, the radar is effectively looking down on top of the aircraft rather than at a very low angled Microwave radar signal head on to the target. This creates a larger return and using Doppler frequency shift is able to establish whether the track is moving, and at what speed. An early heads-up of a potential stealth bomber attack on Russia gives them the advantage of knowing where to send intercept aircraft and set up other defence methods. In the case of an ICBM strike, extra vital minutes warning can be provided. But, as previously mentioned, the position isn’t 100% accurate and can only provide an approximate location of the target – the system can not be used for any weapons fire control.
As previously mentioned, in general Konteyner uses a 50 pps radar signal sent as frequency modulation on pulse (FMOP) using an approximate 12 to 14 kHz of bandwidth. However, through analysis of the Konteyner signals other pps rates of 25 and 100 have been recorded giving ranges up to 6000 km and 1000 km respectively. The manufacturer of Konteyner, NPK NIIDAR (Scientific and Research Institute for Long-Distance Radio Communications), has confirmed the 3000 km range, along with an altitude coverage of 100 km.
One find in my analysis of Konteyner signals was quite interesting.
Quite often when analysing OTHR signals closely, you can see weak Back-scatter return signals between the main pulses. These weak signals travel in the same radar sweep direction as the transmitted ones in either a down-sweep mode from a high frequency to a low one, or in an up-sweep mode from to low to high.
In the image below though you can see another, weaker, radar pulse emanating from the point the first down-sweep pulse ends, travelling up in frequency range and creating a V. If you look closely you can also see a very weak back-scatter signal from both.
My conclusion from this is that the up-sweep pulse is from the 095 degree Konteyner transmitter array, whilst the stronger down-sweep one is from the 275 degree array – the stronger signal is in theory pointing at my antenna in the UK and hence would be emanating from the 275 degree array.
The fact that this signal comes from the 095/275 arrays is a guess of course but I think I’m right. I am also going to guess that the complete radar pulse for the 095/275 transmitters starts at one end of one array, travelling along the 44 masts. When this pulse ends the other array starts in the opposite direction. Moreover, with this method there should be zero interference between the two arrays as they wont be transmitting at the same time.
In the image above, taken from from a screen grab of Procitec’s go2DECODE, you can see that each pulse is every 25 ms, therefore confirming a rate of 40 pps – the software also determines this automatically as shown in the table to the right. Also of note is the analysed signal in the frequency window (Hz) at the bottom. Here you can clearly see the V created by the two pulses.
When we look at the Time display window in go2DECODE (shown below) we can see that I’ve measured the total length of both pulses to be around 6.5 ms. But on closer inspection I think I’ve cut that short a little and it should be 8 ms. This would mean each pulse lasts 4 ms and ties in nicely with the 25 ms per pulse gap as there’s a 21 ms spacing between the end and start of each individual pulse.
I also wonder, that with a gap of 17 ms between the end of the second pulse and the beginning of the first one again, in theory there’s enough of a gap to fit two more 4 ms pulses between these from the the two remaining Konteyner arrays transmitting at 40 pps. Even at a higher 50 pps rate, the 12 ms gap is enough to allow the two remaining pulses to take place with a 4 ms buffer.
This then means that all four Konteyner transmitter arrays can be operational at the same time without causing any potential interference to each other, whether they use the same frequency or different ones. In this case, I’ve been lucky to capture two of the arrays using the same frequency – well, I think I have 🙂
Nevertheless, monitoring the Konteyner signals should bring some further interesting finds, especially if they are using the same frequency occasionally for different surveillance areas. Moreover, it would also be interesting to find all the various pps rates so that system ranges can be established.
Whilst for many, OTHR signals are a pain, wiping out other signals, they still have a lot to give when it comes to SIGINT gathering.
And it may not end at just the one Konteyner system. On the 1st December 2019 it was also announced that a further system would be activated to cover the Arctic region. At the moment, any potential sites have not been mentioned or found, but a likely site would be near Severodvinsk in the Arkhangelsk Oblast, or near Severomorsk in the Murmansk Oblast. Both of these are close to the 1st Air Defence Division headquarters located in Murmansk. My only negative thoughts on this would be that these sites are too close to areas of interest because of the ionosphere skip created, and also probably too far north – ionospheric bounce is not so good towards the poles.
As the original Konteyner transmitter site seems to be being maintained still, be it without any antennas, it also has the interesting aspect of being around 900 km south of the White Sea and areas of coverage needed – perfect for the ionospheric skip. Could this site be changed in aspect so that a transmitter array points to the north to cover the White Sea, Barents Sea and the northern Island? There’s certainly enough room to do this at the Gorodets site.
There has also been mention of another Konteyner site already in construction in the far east. At this time nothing has been found of any construction site that looks to be a Konteyner OTHR and I have my doubts about this. It was first muted in 2010, then again in 2018, and I would have expected something to be there by now.
I highly suspect that this plan has been abandoned, and the 095 degree OTHR of the Kovylkino Konteyner site has taken over the far east coverage.