[Discuss-gnuradio] USRP2, WBX, and Mode S/ADS-B

[Eric Cottrell]

Hello,



I recently got the WBX daughterboard and updated GNURadio after a year or two 
of hiatus.  I really like the WBX so far, and have some ideas for receivers, 
like an ACARS decoder that decodes multiple frequencies simultaneously, or a 
trunked radio decoder that can decode the control channel and one or two voice 
frequencies.  A suggestion would be a WBR daughter board that is receive only.

I updated my Mode S / ADS-B code to work with the latest version of GNURadio 
and the USRP2.  The project now builds outside the GNURadio project.  The USRP2 
allows me to have a 10M sample rate receiver.  I have to go back and update the 
USRP1 version.



I like the updated scope sink.  It is easier to use and I was able to find some 
bugs, including a few stupid bugs, in my code.  It was interesting using the 
scope with various antenna and wideband preamp combinations.  The preamp with a 
1/4 wave antenna worked good.  A colinear decreased range because of a raised 
noise floor and increased interference.  I moved the colinear to my roof 
without the preamp and get a maximum range from 120 to 210 miles depending on 
direction.  I had it up 25 feet off the roof until a rainstorm and wind bent 
the bottom mast section and now it is 15 feet off the roof.  This cut down the 
range a little.  I have a DPD Productions 1090 MHz antenna on order that should 
really fill out the range.


My code implements most of the RTCA recommendations in Appendix I of DO-260B.  
I thought I had a good receiver, but as I fix bugs in my code the receiver 
performance is improving.
http://www.faa.gov/about/office%5Forg/headquarters%5Foffices/ato/service%5Funits/nextgen/research%5Ftech%5Fdev/eng%5Fdev%5Fserv/adsb/wg3/Meetings31/media/pdf/1090-WP31-04.pdf

I only know some basic python so I have external C++ programs that decode the 
log files rather than doing the decoder in Python.  It is an advantage when I 
spot a decoding error.  I fix it and run the same log file through the decoder. 
 I am looking at implementing a socket interface so my code will work with 
Planeplotter and other programs.

I want to try using a matched filter for preamble detection and a method that 
packages the raw samples in a structure when it detects a preamble.  The 
following blocks work on this structure instead of a stream.

There are now several hardware platforms for decoding Mode S and ADS-B, mostly 
from Europe.  The SBS-1 implements a complete receiver/decoder using a FPGA and 
possibly a soft microprocessor.  The RadarBox sends the raw bits over the USB 
and decoding is done on the PC.  These are the major players.  I was surprised 
to see at least three other devices.
http://www.radargadgets.com/
http://www.auroraeurotech.com/avionics/
http://rxcontrol.free.fr/PicADSB/index.html

This seems like such a niche market and yet there are at least five hardware 
solutions and two GNURadio solutions, including mine.  I would like to compare 
my code with a couple of other receivers to see how good it is.

73 Eric