TERN's approach is effective - it works, customers buy it, some people
say its great. However, there is an argument that says something
like the gain you get with S-CDMA is small and not worth the extra
complexity over simpler QPSK with good forward error correction. Now,
there are sides to that argument. My point is that cable plants
have to have clean enough upstream spectrum to keep broadband
interference low, which in turn reduces narrow band, usually therefore
providing some (maybe not a lot of) clean spectrum to run one or
more QPSK channels. That starts the cash flow going. Noise is always
a serious problem when it is out of hand, but can always be managed.
There are many variables in the equation here. Age of cable plant,
type of equipment installed, upstream capability, overall architecture
and topology of the plant, competence of the engineering staff (hate
to say it, but true), topology and geography specific effects,
temperature, and some more. It's not possible to have one simple
solution to everything as there is a lot of grey area. Oh yeah,
I forgot to mention $'s, business plans, $'s in, $'s out, growth
plans, etc.
Mark
|
