First, I am not a deep theory technologist ...
but It looks like you could find some of your answers
in the What is OFDM ? , paper I referenced.
Sure, we cannot re-write the laws of physics, and lower
frequency waves have more energy to lose to physical objects
before they are totally spent.
And refraction (or re-radiation of induced energy) from
solid objects, varies with frequency as well.
All versions ... 11a / g and 16a make use of W-OFDM's
Training Signal technique, to overcome multipath interference ... (not used in DSSS or 11b).
In doing so, they take full advantage of the fact that
in OFDM .... EACH individual SUB-CARRIER has MULTIPATH
Characteristics that are UNIQUE to that INDIVIDUAL Sub-Carrier.
Again, the "Training Signal Technique" makes it all work, according to the Tony Nguyen.
According my interpretation of what the W-OFDM designer told me ...
11a (and thus 11g) was designed (or cobbled
together) WITHOUT a full knowledge of the techniques
described in Tony Nguyen's paper....
and overly large
guard bands were used to compensate for design deficiencies.
I don't have the knowledge to go deeper, but it is
quite obvious that 16a (WiMAX) uses 4 times as many Sub-Carriers,
and in this case MORE IS BETTER, considering
that EACH Sub-Carrier is an UNIQUE Entity, with its OWN
Multipath Interference problems...
More successful transmissions result in more recovered data,
if not the possibility of avoiding multipath interference
(on an individual basis) , in the first place.
And no doubt greater attention was paid to the original
PHYsical Layer (W-OFDM) design (submitted by Wi-Lan to the IEEE) .... as it was used the basis for the 16a PHYsical
Layer.
Current W-OFDM (16a Like) equipment uses a PHYsical Layer
that is about 80 percent identical to the 16a PHYsical Layer, according to info from Wi-Lan.
According to my source again, Bit/Error Rates for
the 16a versions are far better (than for 11a), resulting
in far fewer re-transmissions (under the control of TCP),
and overall greater data throughput results.
3G obviously is a different animal, which uses CDMA,
and the lower frequencies obviously are an asset, in
some cases ..... but not in the case of overall data
throughput, I suspect.
There is NO such thing as COMPLETE Non Line Of Sight,
obviously as we can not penetrate many solid objects ...
so the next tactic is to
COMBINE 256 FFT W-OFDM with MESH or (Wi-Lan's VINE) Technology ..... to go AROUND the solid objects ...
another interesting subject.
BR,
OM |
