Hi Frank:
Concerning the reference infrastructure for
broadband wireless, it would involve fiber to
the cell node, and a relatively fine sectorization
of each cell. The Cellular Vision scheme did not
really include sectorization. The first generation
HP LMDS system had 4 sectors per cell. I think
newer systems will be able to do better. The 2
key elements delaying broadband wireless deployment for
residential customers in 20GHz and beyond range
are: a) the need for inexpensive repeaters to
get around the line of sight requirement, and b)
the need to develop wireless ATM or IP over wireless
software. This software will allow dynamic BW
allocation. Several vendors have already working
software (Stanford Telecom, Hughes Network, and
soon Ensemble Communications for IP over wireless).
However, the standard setting process is just starting.
On the repeater front, I am less aware of new developments.
The key here is that while broadband wireless can be
delivered inexpensively to businesses (large and medium
size office buildings), the economics are not as favorable
for the residential market. So, current broadband wireless
CLECS (WCII, TGNT, and soon NXLK) are going after the
low hanging fruit, and leaving the residential market
for later, with the hope that deployment costs will come
down. Waiting too long may be a mistake-- for example,
all of the apartment buildings in NY to which RCNC is
bringing fiber could have been reached more cheaply with
wireless equipment.
Best regards,
Bernard Levy |
