Mode S Technology
Interview with Bill Harman - 11/3/00
Worked at trw. worked on the pioneer jupiter spacecraft.
Decision was made at the mit level to diversify and get involved in a
large number of faa programs.
100 responses/sec transponders. u.s. is 50-70/second. not unusual to
reach 100 though. in frankfurt it's about 200 respones/sec. if you
multipled all the signals end-to-end they would use up the entire
timeline. the system was successful under those conditions.
worked well for main test cases there were some cases where it would
reach 50 miles but in some cases it was 20-30 miles. trying to figure
out whether we did something wrong in the testing or this is a real
limitation in the system. should work for all aircraft. so the
differences were different aircraft. the criterion for successful is
something voted in arbitrarily.
vdl mode 4: it is better in the frequency band in certain
respects. it's a lower frequency so the signals propagate better so
that it bends a little better. it would be significant on an airport
surface so fewer stations would need to be installed around logan
airport, for instance. gps squitter at logan 4-6 stations, but the vdl
system would require 2. covering airplanes all the way in to the
airport would be better. the bandwidth is considerably less so it is
hard to be sure that there is enough to be sure there is enough for
all the aircraft. it is a dedicated system though.
mode S should it be on the same frequency band (DABS)? that was the
original debate. far easier for the avionics for the system. both the
interrogation and reply were free to chose any modulation. the first
choice was to stay on the same frequency. we ended up making this
DPSK. began by listing the choices.
"We have to go from where we are now into the future and can't just
stop everything and have the new one take over. Aviation has to
continue on things that are on the airplane really need to be paid
attention to." Legacy now has a negative connation but used to have a
positive connotation. The concept of taking the existing Mode S
transponders (NOT legacy) because it is important to be compatible. it
is better to keep it if we don't have to throw it out.
some of the bugs: found gaps in reception. they were due to the data
the encoding of latitude and longitude is data compression 17 bits. if
it were direct binary it would be 24 bits. it turns out that the
compression algorithm missed some of the end point boundary
conditions. so various minor bugs that need to be worked out in the
design and implementation.