Phil, you probably know an order of magnitude better than me, but anyway: if you have the freedom to design the transmission mode, have a look at the techniques used by DVB-T/DVB-H, which face similar or worse multipath problems (transmitter is much closer to the ground): use COFDM, splitting the bandwidth into multiple subcarriers, and them modulate each subcarrier using QPSK or even QAM. Then you are making the symbol period much longer, improving immunity against multipath; but at the same time your symbols are much more complex, thus binary efficiency is not much worse than a similar single-carrier system. At high speeds doppler becomes a problem, but it's a matter of achieving a balance between multipath immunity and doppler tolerance.
Miguel EA1ICZ 2013/7/17 Phil Karn <k...@ka9q.net> > I'm one of several adult mentors to a local high school ham club that's > been designing and flying balloons. We are designing a high speed 70cm > digital downlink for a future payload. One of my major concerns is > frequency selective multipath fading in the 1 MHz bandwidth I plan to > use as the balloon gets low on the horizon of the tracking station and > the receiver begins to pick up a reflection off the ground. > > I think one of the simpler ways to deal with multipath is to use > circular polarization on both ends. This automatically rejects all > odd-order reflections (including the all-important 1st order reflection > off the ground) as they have the opposite circularity sense to the > direct signal and automatically be cancelled at the receiver. (CP was > tried experimentally with analog TV broadcasting back in the 1970s as a > means of reducing ghosting.) > > On the ground we'll just use a regular 70cm satellite antenna, but on > the balloon I need a 70cm circularly polarized transmit antenna with a > fairly wide (hemispherical) beamwidth that can maintain its circularity > over as much of that as possible. > > I'm not an antenna guy, but my understanding tells me that a quadrifilar > might be a good choice here. Any comments on this? Can anybody point me > to some typical performance figures for these antennas? Again, I'm more > interested in cross-polarization rejection than in absolute gain. I have > plenty of link margin and I'm willing to sacrifice a few dB if necessary > to get rid of those deep multipath fades. > > Thanks, > > Phil > > _______________________________________________ > Sent via AMSAT-BB@amsat.org. Opinions expressed are those of the author. > Not an AMSAT-NA member? Join now to support the amateur satellite program! > Subscription settings: http://amsat.org/mailman/listinfo/amsat-bb > _______________________________________________ Sent via AMSAT-BB@amsat.org. Opinions expressed are those of the author. Not an AMSAT-NA member? Join now to support the amateur satellite program! Subscription settings: http://amsat.org/mailman/listinfo/amsat-bb
