[UFO Chicago] Question: wireless internet access in Chicago?

[Neil R. Ormos]

On Thu, 30 Dec 2004, sjk wrote:
> On Thu, 30 Dec 2004, Neil R. Ormos wrote:
>> On Wed, 29 Dec 2004, sjk wrote:

>>> JMHO, AWB is an example of what's wrong with a
>>> number of wisps - they flood areas with canopy
>>> signaling - creating more pollution, they have
>>> extremely high latency, and their network is not
>>> well managed.  With the deployment of WiMax
>>> quality wireless offerings may appear next year
>>> in urban areas, but till then I'd suggest you
>>> avoid any mass market wisp.

>> I don't know anything about AWB, but what are you
>> suggesting a wireless ISP is supposed to do other
>> than flood their service area?  And how will WiMax
>> be any less pollutive than Canopy or other
>> existing wireless solutions?

> WiMax is more of a cell based technology
> allowing for multiple IDs and channel power. The
> spectrum used by almost all point to multi-point
> wireless systems (2.4GHz) is extremely polluted
> -- by flooding large areas, instead of targeting
> specific service areas, you create interferance
> problems. In wifi -- for example -- you only
> have 3 usable, non-overlapping, channels
> available. As channels congestion increases the
> overall quality of service decreases. Within
> large urban areas the congestion is so bad that
> the networks can become unusable.

> The Canopy system, employeed by AWB, works as a
> quasi-mesh architecture; however, as nodes are
> added the overall amount of RF signaling causes
> so much contention that network performance
> suffers. Instead of flooding an entire area with
> RF, it sould be better to target service
> locations with directional sector antennas.

It may well be better from a spectrum management
point of view to serve users via smaller service
areas, although whether that's done by using
directional antennas or other means is entirely
irrelevent.

However, I still don't understand why you
deprecate AWB.  It's not clear that smaller
service areas would have been economically
feasible using the technology available when that
system was deployed.  It's also not clear that
WiMax will be any better, once it is deployed on a
large scale.  WiMax is said to have a maximum
range of 30 miles, although the optimum range is
less than 10 miles.  Network operators are likely
to try to serve customers using the largest
practical service areas from a base station, at
least until demand requires otherwise.  Those who
don't will go out of business a la Ricochet.
WiMax base stations are expected to be less
expensive than Canopy base stations, but backhaul
remains an expensive problem.

It's also not clear that WiMax, as actually
deployed, will be a desireable broadband
alternative for those, like Brian, who are unhappy
with 300 Kbps throughput.  An article by Intel
claims that "[s]ervice delivery to end clients is
likely to be roughly 300Kbps for residences and
2Mbps for businesses."  That's better than dialup,
but not exactly spectactular.

As for "pollution," as you appear to be using the
term (in its non-traditional sense), WiMax is
likely going to make the situation worse, because
its low cost will stimulate new demand.  A substantial
fraction of the spectrum is expected to be
unlicensed; that coupled with the lower cost of
the base stations is likely to invite large
numbers of unqualified operators.  One apparent
spectrum management benefit offered by WiMax is
that bandwidth is positively allocated to
subscribers by the base station in a
time-muliplexed fashion, reducing the congestive
effect of contention when demand is high.  It
remains to be seen whether this feature offers any
protection to the noise floor once real systems
have been deployed.

--neil