Menu Sign In Contact FAQ
Banner
Welcome to our forums

Software Defined Radio

I’ve recently been playing with a little TV tuner dongle which, with some open source drivers, can be made to work as a pretty interesting little SDR.

It makes me wonder why they haven’t taken over the radio avionics market. A slightly more expensive SDR can cover the whole of the aviation bands easily (you can buy SDRs with 30Mhz of bandwidth for less than £300). You could theoretically monitor every VOR within range and use it do VOR – VOR positioning etc.

The manufacturers probably don’t see much of a point in it when you have a WAAS GPS, which will always be more accurate than VOR-VOR positioning.

Cheap SDRs (e.g. the ones you can buy for £300) tend to have a problem with deaf receivers and relatively poor dynamic range and high power consumption and require a fairly powerful computer to process several MHz of bandwidth at once (a Raspberry Pi 4 for instance, which is about the limit of what you’ll be able to cram into a half height radio tray, can only manage to handle about 2MHz with the LimeSDR – you need a full desktop computer to do 10MHz of bandwidth) – it’s easier and cheaper to make a traditional receiver which is sensitive and has good dynamic range (you can probably make a better traditional receiver for a tenth of the cost).

Maybe some time in the future it’ll be possible (perhaps someone will make an ASIC that can do it without needing to ask too much of the box’s general purpose CPU) but you’ll still have the question: why, when you have a perfectly good WAAS GPS, with chipsets you can just buy off the shelf, and even more VORs have been decomissioned?

Last Edited by alioth at 20 Dec 10:02
Andreas IOM

Yeah, I think those are very fair points. For the most part it is probably just a case of the technology maturing at the wrong time. As you say, conventional ground based navigation is in its twilight years.

I’m sure on those rare occasions GPS is being jammed it would be nice to have the aircraft navigation systems working relatively normally but that hardly makes a big selling point.

LondonMike wrote:

I’m sure on those rare occasions GPS is being jammed it would be nice to have the aircraft navigation systems working relatively normally but that hardly makes a big selling point.

It does make a selling point, and the need for a jam-proof navigation system has been repeatedly mentioned at many levels. However, VOR-VOR doesn’t really get much consideration, probably because of its low accuracy far away from the beacon (noticeably worse than DME-DME). On the other hand, the more VORs you take into account, the more accurate it gets, and your idea of monitoring the entire frequency range is good because it derives extra benefits from the existing ground infrastructure (while it lasts).

LKBU (near Prague), Czech Republic

We have previously discussed why nobody has made a replacement for the widely used KX155* radios – e.g. here. And if someone did, they may well have chosen to do it using DSP techniques i.e. mix down to say 10MHz (direct processing of the RF would need sampling to maybe 16 bits at a few hundred msps and while that exists for e.g. Lidar it is expensive and the really good chips are export restricted, plus you need awesome processing hardware) or so and then use a fast A-D converter and do the whole receiver in software. The problem is that GA avionics firms are very short of real design expertise, not to mention “vision”.

One reason might be that the transmitter still has to be done the “old way” and all radios since the 1970s or so use PLL synthesis anyway. And in a NAV/COM box you have to receive the LOC/VOR and GS at the same time (so a KX radio which does ILS contains three separate receivers) so while the DSP method could receive all of these at the same time (a frequency hopping receiver) that needs even more performance.

You should not need a PC-level CPU if you use some FPGAs.

Re building a backup for GPS, the way to do that would indeed be DME-DME. That is what airliners use to fix up INS errors (although the recent ones can also use GPS). VORs may over many years get removed but DMEs will not be because along with ATC radar they provide the “final resort” for airliners in case of widespread loss of GPS. I went to a Eurocontrol “nav workshop” in 2008 where this was covered and the plan then was to definitely not reduce DME coverage. In fact France has installed many DMEs in their VORs most of which previously had no colocated DMEs. I still have some notes from that workshop; I went there on behalf of a UK pilot group.

The problem is that GPS works so well, there is just no market for such a device. It has been tried, maybe 20 years ago, and failed. It was Collins or Narco or some such who made an “RNAV box” but nobody bought it.

Administrator
Shoreham EGKA, United Kingdom

Ultranomad wrote:

On the other hand, the more VORs you take into account, the more accurate it gets,

Not that simple…

While it is good to have a good choice of VORs tho chose the two best from, it’s actually not trivial to actually use more than 2 of them at the same time to get a better positioning accuracy.

Germany

Peter wrote:

direct processing of the RF would need sampling to maybe 16 bits at a few hundred msps and while that exists

Pretty much all the SDRs you can get for reasonable prices do the sampling of an IF (intermediate frequency), not direct – and any reasonable cost SDR generally has an 8 or 12 bit ADC (the much more expensive Ettus USRP, which costs over £1000, has a 14 bit ADC). You can do direct sampling for HF frequencies (up to about 30 MHz) without too much of a problem, but once you’re looking at 100+ MHz it’s just so much more cost effective to have the ADC and signal processing working at some lower IF.

Andreas IOM

Malibuflyer wrote:

While it is good to have a good choice of VORs tho chose the two best from, it’s actually not trivial to actually use more than 2 of them at the same time to get a better positioning accuracy.

The only ones that are potentially better are the closest ones, otherwise there is no a priori rule. So the first stage in any event would be to perform multiangulation on all VORs. Then, if you wish, you can do the second approximation: use the position obtained to estimate the distance to individual VORs, then either drop the remote ones altogether, or do weighted multiangulation with each VOR’s weight inversely proportional to the distance from ownship.

LKBU (near Prague), Czech Republic

Ultranomad wrote:

The only ones that are potentially better are the closest ones, otherwise there is no a priori rule.

Choosing the two best ones is an optimization between distance and relative bearing (which can be easily seen by the fact that with two VORs that are on the same straight line with the plane you can not determine a position at all).

If you have three VORs available one can show that the gain in accuracy by using all three compared to just using the best 2 of them is less than 5%. Adding more VORs further reduces this accuracy gain per VOR.

Germany

Malibuflyer wrote:

If you have three VORs available one can show that the gain in accuracy by using all three compared to just using the best 2 of them is less than 5%.

5% of what?

ESKC (Uppsala/Sundbro), Sweden
11 Posts
Sign in to add your message

Back to Top