Dear Alberto, LF Group,
With OFDM being a fashionable form of modulation for all sorts of things
(DAB, DTTV, DRM, next generation wireless widgets of all shapes and sizes),
there is a huge amount of interest in reducing the crest factor so that PAs
can be made more efficient, or less linear - if you do an internet search
with words like ofdm, cofdm, peak, mean, power, crest, factor, reed, muller,
pmepr you will find huge amounts of info on this topic - probably more than
you wanted! In a cursory search, I did not find any useful,
easy-to-understand articles, but that does not mean there aren't any. A few
of the professional journal papers I skimmed through were:
http://www.kmutt.ac.th/itc2002/CD/pdf/19_07_45/FA2_OD/2.pdf
http://www.rennes.supelec.fr/ren/rd/scee/ftp/articles/melecon2000.pdf
http://www.rennes.supelec.fr/ren/rd/scee/ftp/articles/cscc2000.pdf
These are ally highly abstract, and confirmed that I know almost nothing
about the subject, but do appear to say that peak-to-mean power ratio can be
reduced to 3dB, or less, by clever coding. Most systems using OFDM seem to
limit the crest factor in one way or another, so I think it is likely
suitable techniques already exist out there, somewhere... how applicable
they are to LF is another matter, of course.
An intriguing idea is the "single frequency network" capability of COFDM,
where the same signal can be transmitted by multiple transmitters, and
received without suffering greatly from selective fading type problems that
would happen with most signals. This reminds me of the talk there was a
while back of using multiple LF transmitters for "beam forming", and
multiple receivers for "diversity".
Cheers, Jim Moritz
73 de M0BMU
|
