[CUWiN-Dev] G and A and realworld throughput

[listsubs0506]

A couple of interesting articles re: throughput in multi hop wireless 
networks:

http://www.compliancepipeline.com/159905154

http://black.csl.uiuc.edu/~prkumar/ps_files/exp.pdf

I see that a co-author of the second article (which is cited by the 
first) is P.R. Kumar of the University of Illinois Coordinated Science 
Laboratory, 1308 West Main Street, Urbana IL  <prkumar at uiuc.edu>.

By the way, at risk of going a little off-topic, those articles were 
cited in footnotes 27 and 28 of a Request for Reconsideration recently 
filed with the FCC by Intel Corporation / Redline Communications, Inc. / 
Alvarion, Inc.

They are asking the FCC to reverse its recent decision that would enable 
community wireless and pretty much everyone else to have access to 3650 
- 3700 MHz spectrum in urban areas such as Albuquerque, Atlanta, Austin, 
Boston, Chicago, Cleveland, Columbus, Detroit, Houston, Kansas City, Las 
Vegas, Milwaukee, New Orleans, Oklahoma City, Phoenix, San Antonio, and 
Tucson. That particular Request doesn't ask the FCC to reverse its grant 
of widespread access in less heavily populated areas areas, though other 
recent requests for reconsideration do ask for half of the band even in 
rural areas to be granted to exclusive licensees.

Intel et al state:

"Petitioners believe that exclusive licensing in the Top 50 MSAs will 
promote optimal quality of service ("QoS") and strong business 
investment certainty in these markets; such results are not possible 
with self-coordinated contention protocols based on the mutual 
obligaiton to cooperate."

"Specifically, Petitioners believe that two blocks of 25 megahertz each 
should be exclusively licensed in these markets."

In other words, Stelios, you would almost certainly be locked out in 
Chicago. And the community wireless folks here in the Boston area would 
also almost certainly be locked out if the Intel Request for 
Reconsideration is successful (I imagine in theory, the city could be 
given one of those licenses and then permitted to sublease... but I 
don't think the petitions for reconsideration mention possibilities 
other than assignment by auction). 

I'd be curious whether any of you are in touch with P.R. Kumar and able 
to determine whether he thinks his research has been properly used in 
support of the arguments presented in the Petition. Some of those 
arguments are remarkably dismissive of Wi-Fi and other systems that 
might be used in the band, e.g., footnote 22:

"In this regard, Petitioners note that Wi-Fi works because approximately 
a dozen users (i.e., a very limited number of users) are sharing 90 MHz 
of spectrum. In contrast, the FCC's new rules for the 3650 MHz band 
would allow dozens, or even hundreds, of users to share a mere 50 MHz. 
That many users trying to access this limited  amount of spectrum would 
be detrimental to transmission capacity -- causing it to degenerate to 
something less than broadband or even no transmission at all."

The full Intel Request for Reconsideration is available at:
http://makeashorterlink.com/?N2F612D6B

or you can find the full set of about 9 Requests for Reconsideration in 
that FCC Proceeding by entering 04-151 in the top left hand corner 
search at www.fcc.gov... I'd be glad to send summaries/extracts of them 
to anyone who's interested but doesn't have time to plow through them all.

  - Steve Ronan