24 Jan
2022
24 Jan
'22
4:05 a.m.
We have been considering doing that kind of thing to triangulate abusers on our multi-site
repeater PI2NOS.
At some point in time it had 25 receivers and 5 co-channel transmitters covering the
entire country,
so there were always miscreants active.
Some work has been done, e.g. the general location was determined by comparing the
relative
receive signal strength on the 25 receiver sites, but that only worked for strong signals
that
hit multiple receivers, and were themselves transmitting omnidirectionally.
The typical mobile or portable user could not really be located.
With TDOA the problem is that you cannot really timestamp a signal with such a small
bandwidth
accurately enough. It requires correlation of the signal and that does not work too well
on a
barely modulated NBFM carrier. A German amateur demonstrated use with DMR and there it
is much easier because there is a nicely structured and clocked signal even when the user
does
not talk. But to really get it to work you need wideband signals. There is a reason why
GPS
systems use spread spectrum!
Of course with doppler direction finding it is possible to locate all signals, also NBFM
or unmodulated
carriers, but it requires a dedicated antenna which would make it more difficult to
deploy.
I recently saw an announcement for a board with I think 5 circuits similar to the RTL-SDR
clocked
from a common clock source, and it should be possible to do direction finding with that.
(the classical solution uses a single receiver and a pin diode switch to connect the
antennas
alternately. the receiver ideally should have both FM and AM detection)
With this solution you do not need accurate time.
Rob
On 1/24/22 09:13, Matt - VK2FLY via 44Net wrote:
I did the math on RDF using 2 and 70 using difference
time of arrival. The math worked out that I needed gps clocks to get accuracy down to a
50m with 4 receiving stations. However the gps clocks increased the price somewhat. I did
not investigate the latency involved with RTL dongles or even if that latency could be
fixed/measured which would be a key component. A dedicated radio may also need to be built
to get RX carrier detection latency predictable. But in theory it would be possible to
build a network with unlimited cheap RX stations reporting locations on a map APRS style
from every received signal.
Matt
On 24/1/2022 5:11 pm, Tony Langdon via 44Net wrote:
> On 24/1/22 5:48 pm, Steve L via 44Net wrote:
>> On Sat, Jan 22, 2022 at 11:14 AM Robert Simmons via 44Net
>> <44net(a)mailman.ampr.org> wrote:
>>
>>> Does anyone here have any interest in remote reporting radio direction
>>> finders ? ( RDFs )
>>>
>> I think if it can be developed as open source software using cheap
>> hardware like the RTL-SDR (the kiwisdr is expensive), that would be a
>> very good project worthy of an ARDC grant.
> I agree. An RTL-SDR may or may not cut it (one way to find out ;) ),
> but open source software, and open and inexpensive hardware would be
> interesting to pursue.
>