Trasponder C mode

[Admin]

What sort of range can you detect targets out to?

This is the black art I am dealing with at the moment.
There is no range in the message, simply signal strength and altitude.
The idea is to convert signal strength to range, to be honest I think this is fraught with inaccuracies, but I am willing to investigate.

Thx
Lee

[SteveN]

The OGN people show a method for calibrating dongles which I guess would be needed to do range estimation based on signal strength:

http://wiki.glidernet.org/wiki:raspberry-pi-installation

The very first thing you need in order to receive the signals is to be on the correct radio frequency.
Cheap SDR receivers use cheap crystals and their frequency tolerance is about +/-50-100ppm.
At 868MHz 100ppm error makes you 86.8kHz away from the correct frequency, while frequency deviation for FLARM signals is +/-50kHz.
Technically, the software receiver can search a wide range of frequencies for radio packets but this inflicts lot of CPU, thus can only be done for stronger CPU boards, but not for Raspberry PI.
Thus you need to know how much off is the crystal of your DBV-T dongle before you proceed.

You can measure the crystal with the gsm_scan tool, run it like this:

./gsm_scan --ppm 50 --gain 20

it should receive some GSM broadcast channels and measure the frequency correction.
Notice the GSM frequency with a strong broadcast channel, even better: with one or two directly adjacent channels.
Adjust the gain (--gain) and initial crystal correction (--ppm) for best reception of as many as possible channels with consistent correction measurement.
GSM signals are very strong, thus too much gain is not good. You need to find an optimal setting here.

.
I've tinkered with this on my spare pi using my 2 dongles.  One is 15ppm out, the other 30ppm

[rg]

Well if you could detect mode c at say 3, 1 & 0.5nm  (of course accepting it's may not be accurate ) that would be incredibly valuable.

[The Westmorland Flyer]

The idea is to convert signal strength to range, to be honest I think this is fraught with inaccuracies, but I am willing to investigate.

Completely agree. It will be nearly impossible to determine range using signal strength with even the slightest nod to accuracy. Amongst the many problems is the fact that the transponder antenna is on the underside of the aircraft - the obvious place, since it is working a ground radar. Signals above or even along the plane of the aircraft will be severely attenuated compared with those below the plane, making range determination on signal strength alone nearly impossible. Without directional antennas it isn't possible to determine the bearing either. I don't think it's worth the effort.

[Admin]

sorry let me clarify - it is the vertical height difference I am interested in, eg

+/- 250ft RED
+/- 500ft AMBER
+/- 1000ft GREEN

The diameter of the circle is the indication of range (derived from signal strength)

[Admin]

The idea is to convert signal strength to range, to be honest I think this is fraught with inaccuracies, but I am willing to investigate.

Completely agree. It will be nearly impossible to determine range using signal strength with even the slightest nod to accuracy. Amongst the many problems is the fact that the transponder antenna is on the underside of the aircraft - the obvious place, since it is working a ground radar. Signals above or even along the plane of the aircraft will be severely attenuated compared with those below the plane, making range determination on signal strength alone nearly impossible. Without directional antennas it isn't possible to determine the bearing either. I don't think it's worth the effort.

The trick I am hoping to use is as follows.
Gather and store the ADS-B messages as references,
this will give me an indication of strength for relative distance/height.
this provides me a look up table averaging strength against the distance/height.
on reception of a non ADS-B message, index the table of known (ADS-B) messages for the best estimate.

So in effect the receiver becomes 'self calibrating' using the well defined ADS-B data, to estimate the non ADS-B data

Determining bearing is totally impossible, that is why I said 'bearingless traffic', although I may well be able to use
signal value increasing/decreasing.

I think this is worth a shot, please keep to youselves, I am still waiting for the patent grant on the self calibrating receiver idea 

Thx
Lee

[SteveN]

Clever Clever Clever

[rg]

sorry let me clarify - it is the vertical height difference I am interested in, eg

+/- 250ft RED
+/- 500ft AMBER
+/- 1000ft GREEN

The diameter of the circle is the indication of range (derived from signal strength)

Sounds good also. 

[The Westmorland Flyer]

Yes, that is indeed a cunning plan and it will be very interesting to see how it works out in practice.

[Keithvinning]

And where is this in the hierarchy of priorities 

[scsirob]

The trick I am hoping to use is as follows.
Gather and store the ADS-B messages as references,
this will give me an indication of strength for relative distance/height.
this provides me a look up table averaging strength against the distance/height.
on reception of a non ADS-B message, index the table of known (ADS-B) messages for the best estimate.

I don't want to rain on your parade and it is indeed a smart approach, but its success hinges on a chicken-and-egg problem. At this time the ADS-B traffic is mostly from airliners at 30000+ ft. Those use transponders with significantly higher output than most GA planes (they signal level is much higher than our transponders) and fly where we don't. Using those as a reference may give the false impression that a relatively weak Mode-C is still at safe distance.

The number of usable ADS-B references at VFR / GA flight altitudes is almost non-existent. In fact I am getting phone calls asking why my rinky-dink toy plane shows up on FR24 between the big boys..

The lack of ADS-B equipped GA planes is one of the reasons for PAW has its own ARF transceiver in the first place. If there were enough ADS-B equipped GA planes then you wouldn't need to set up ARF transmitters to announce your location.

[Admin]

Hi Rob

Its a good point, I had not realised there was such a wide variation in the output power on the transponders. So the only thing that can be reported is proximity to vertical separation.

Well if anything at all.

Thx
Lee

[rogerabc]

Mode C deconfliction will be important to encourage mass adoption.
My Zaon PCAS does a very accurate job detecting range and vertical separation.
Can you reverse engineer their system Lee?

[The Westmorland Flyer]

It's certainly true that the transponders in CAT are generally more powerful than those fitted to (low end) GA but I doubt the difference is more than a few dB. Any attempt to compute distance from signal strength will anyway come up against the inverse square law, meaning that closer signals will be considerably stronger. This reduces the impact of power differential though it doesn't of course eliminate it.

The point about few ADS-B aircraft in "our" space" is absolutely correct. As I said earlier, I think in the fullness of time this will change but it needs a low cost ADS-B solution. I think if that were to happen then GA would adopt it because, unlike Mode-S, 8.33kHz and so on, it has an obvious and strong benefit to GA. In that scenario PAW is in an excellent position to become the combined ADS-B out/ADS-B in device of choice for GA.

I'm looking with interest at the current ADS-B trials. Not much use to me up here in Cumbria but a step in the right direction for sure.

[scsirob]

Well if anything at all.

One thing you can use to filter is that ADS-B signals reporting > 250kts are probably not usable as a reference?

I'd be interested to know how many of the people on this forum have Mode-S transponders, how many of those are Extended Squitter capable, and how many actually have that enabled. In The Netherlands, Mode-S is mandatory, but even so I still see only few enabling ES.

Would that be something for a poll on the forum?