Topics

Testing Resources

Paul Johnson <ve7dhm@...>
 

Brian,

The remote location may or may not have internet access.

The general concern for me is how do I prove that the initial
installation is good - the grounding, cabling, connections etc - then
add the RF emitter, and then the RF path - so that when I walk away from
a remote site installation I only have to be concerned about the RF
path.

I would like to see some general usage installation tools which would
provide a "good to go" check list. This means being able to locally
loop back a few megabits of data with no errors and then do the same
end to end over the RF path. Not all installations will be created
equally and it would be nice to be able to identify where data link
errors are injected into the UDR-X network...is it the installation or
the RF path?

I have used a Firebird and Fluke network equipment to do end to end
testing and I thought there must be software applications that can also
do some of the testing done with such equipment checking "keyboard to
antenna".

It is my understanding that the UDR-X is not full duplex so looping back
from RF output to RX input, using a RF signal sampler - Microlab / FXR
type - is not possible. So, the alternative is to use two UDR-X units
which adds additional complexity to the testing procedure but is better
than nothing.
The ability of the modem and emitter to operate full duplex certainly is
the ideal local test procedure without having to "go on the air". Using
two UDR-Xs would accomplish the same goal.

For end to end testing over the RF path here again full duplex would be
ideal. However, something similar could be accomplished by a store and
forward application running at the remote end. The local end would TX a
data stream of N bits and the remote end store and then TX back the bit
stream to the local end. A checksum, SNR, BER, and constellation
display application would give RF path quality.

There seems to be quite a bit of info on the internet regarding python,
ipython, matlablib, SNR, BER and linux. So, if the UDR-X computer is up
to the task then the possibility of general usage installation apps
running on it to test an installation looks promising....or maybe adding
a Raspberry Pi might also work.

Lots of things to think, do and have fun with (;

Looking for ideas....

Paul VE7DHM



Hi Paul,
When you say remote location, I assume you have no network access,
only RF

Looping 25W back into our receiver would be a bad idea. :) Looping the
modem
would just tell you the processor is running.
Do you mean more of a digipeater mode where received packets are
stored then
re-transmitted? We could disable checksum to guarantee a response for
analysis
locally.

As part of the Hi-Speed effort, we are building a channel sounder for
testing
purposes. This is a PN generator designed for analysis.
The easiest thing to do would be to put the remote UDR in sounder mode
and use
your local UDR to analyze the received data which includes RSSI and
BER.
Constellation is an internal design tool at this time, but we will
provide the
hooks to some enterprising WebGL folks for display.
The results could be logged locally. We could also run sounder as a
chron job
(telemetry or beacon) then you'd get results even if you couldn't
successfully
talk to the remote unit. A few second burst hourly would be
interesting.

Great topic. What if you had several internet connected UDRs receiving
results
for comparison?
I'd like more input, I think this topic has a lot of merit.
Bryan - K7UDR

"John D. Hays" <john@...>
 

Hi Paul,

Some very thought provoking ideas.

These are good tools for the community to start thinking about.  To that end, it is our plan for the API to include socket access to the modem, as well as many of the system metrics that would aid such tools.  As a small company, we will not be able to develop every tool that the community will want or need, but will support developers who wish to undertake their creation -- part of our commitment to an open architecture.

The processor is capable of running Python, C, JavaScript/NodeJS, etc. using standard Debian Linux interfaces and tools.

As we prepare for Dayton, our schedules will be pretty busy through the rest of the week.

Watch for some news from Dayton on Friday.  [ Sorry, it will not be that UDRX-440s are shipping :( ]


John D. Hays
K7VE
PO Box 1223, Edmonds, WA 98020-1223 
  





On Tue, May 13, 2014 at 3:26 PM, Paul Johnson ve7dhm@... [UniversalDigitalRadio] <UniversalDigitalRadio@...> wrote:
 

Brian,

The remote location may or may not have internet access.

The general concern for me is how do I prove that the initial
installation is good - the grounding, cabling, connections etc - then
add the RF emitter, and then the RF path - so that when I walk away from
a remote site installation I only have to be concerned about the RF
path.

I would like to see some general usage installation tools which would
provide a "good to go" check list. This means being able to locally
loop back a few megabits of data with no errors and then do the same
end to end over the RF path. Not all installations will be created
equally and it would be nice to be able to identify where data link
errors are injected into the UDR-X network...is it the installation or
the RF path?

I have used a Firebird and Fluke network equipment to do end to end
testing and I thought there must be software applications that can also
do some of the testing done with such equipment checking "keyboard to
antenna".

It is my understanding that the UDR-X is not full duplex so looping back
from RF output to RX input, using a RF signal sampler - Microlab / FXR
type - is not possible. So, the alternative is to use two UDR-X units
which adds additional complexity to the testing procedure but is better
than nothing.
The ability of the modem and emitter to operate full duplex certainly is
the ideal local test procedure without having to "go on the air". Using
two UDR-Xs would accomplish the same goal.

For end to end testing over the RF path here again full duplex would be
ideal. However, something similar could be accomplished by a store and
forward application running at the remote end. The local end would TX a
data stream of N bits and the remote end store and then TX back the bit
stream to the local end. A checksum, SNR, BER, and constellation
display application would give RF path quality.

There seems to be quite a bit of info on the internet regarding python,
ipython, matlablib, SNR, BER and linux. So, if the UDR-X computer is up
to the task then the possibility of general usage installation apps
running on it to test an installation looks promising....or maybe adding
a Raspberry Pi might also work.

Lots of things to think, do and have fun with (;

Looking for ideas....

Paul VE7DHM



Mike Heitmann <n0so@...>
 



Watch for some news from Dayton on Friday.  [ Sorry, it will not be that UDRX-440s are shipping :( ]




Hopefully someone will post the news here for those of us unfortunate enough to not make it to Dayton this year ( Granddaughters First birthday party for this OM ).

Mike, N0SO