more on two more on Sobering analysis of WiMax

[David Farber <dave () farber net>]

------ Forwarded Message
From: "S. Srinivas" <srinivas () radiofiber com>
Date: Mon, 07 Feb 2005 15:05:02 -0800
To: <dave () farber net>, <jim () netgate com>, <jeff () jeffporten com>
Subject: Re: [IP] two more on Sobering analysis of WiMax

Jim, Jeff
 
By way of introduction, I was founder / CTO at CoWave Networks - later
merged with Advanced Radio Cells, now I am completing a 100 Mbps NLOS radio
implementation that is proprietary.
 
I agree with some aspects of your take, but not all. Agreed WiMax has been
overhyped and I always believed and still do that standards that succeed are
not written by a bunch of companies getting together. (Even WiFi started
mainly by the efforts of Lucent (and Harris) I believe.  The more common
scenario is the "creation" of a defacto standard as in the microprocessor
architectures by Intel - or Microsoft's OS/Office. I think if a company
succeeds at creating a low cost ( CPE < 150$, BS < 500$) system - say based
on structured mesh networking, and if the architecture makes it unnecessary
to get tower spaces on typical crowded towers, then there is a good chance
of success. None of this is really part of WiMax today and so there is
plenty of room for proprietary systems still or ones built on top of WiMax.
 
I dont buy the assertion that having wires makes wireless impossible to
succeed. Even wires need maintenance and if you look at DSLAM costs of
installation and all the associated costs of modem and installation or for
that matter of T1s or cable modems, the costs are high enough to allow
wireless to effectively compete even where SBC Yahoo DSL is 26.95 (19.95).
Capacity is not the only issue at work. Coverage and cost of roll out /
maintenance are more relevant. Infact capacity is the last important
parameter. 
 
Most companies owning 3.5 GHz and MMDS are very interested in WiMax - for
Sprint this is a major boon. For big companies this is essential since they
can dictate pricing etc. of the devices and not the other way around like in
cellular. For eqpt vendors this is indeed a death trap. Dont make WiMax eqpt
implies no chance of taking any share of the pie. Making it implies risking
much NRE upfront and possibility of never recouping it due to pricing
pressures. Infact I am lost for words at the folly of so many chip companies
in this space  for fixed wireless.
 
However it is not that clueless to be an unlicensed ISP. There is enough
spectrum in 5 GHz unlicensed to handle upwards of 1 Gbps capacity (perhaps 2
to 4 Gbps with smart antenna techniques). This would work rather well in any
area (with QOS I may add) as long as the network can smartly avoid
interfered regions (aka. cognitive radios). Also the last connectivity can
be in 2.4 GHz or even lower unlicensed as long as the coverage is 1000 ft or
less - small picocells which by analogy with WiFi will avoid interference as
well, however getting greater NLOS coverage through trees to make indoor
distribution possible. I am not that tied into NLOS indoors since startup
ISPs can do just as well with rooftop installations that are easy -
especially internationally.
 
Srinivas
 
 
 
------ Forwarded Message
From: Jim Thompson <jim () netgate com>
Date: Mon, 07 Feb 2005 09:14:37 -1000
To: <dave () farber net>
Cc: Ip <ip () v2 listbox com>, <jeff () jeffporten com>
Subject: Re: [IP] more on Sobering analysis of WiMax

> 1) Yes, WiMax is not going to be available for a while.  The one bit of
> news I hadn't heard was in regards to certification; that having been
> said, if you'll recall the early days of 802.11a and b, you can expect
> to see some providers rolling out early at the risk of not being
> compatible later.

but of course, these types of "solution" are available today, and
haven't made "wireless broadband" a reality.

> 2) A technology that can "plug holes in [rural] broadband coverage" is
> a VERY big deal in countries with large uncovered areas.  The article
> rather condescendingly talks about getting broadband to poorer nations,
> and I'm recalling how difficult it is to get decent access in Ohio.

Repeat after me, "you'll never replace wires with radios".   There will
always be more capacity in
a wire than in a RF link.  Always.

The thing that makes wireless "work" is that it enables mobility.   If
you don't have a mobile aspect to your

Yes, the convenience of not having to locate outlets make WiFi nice in
coffee shops and airports.  I helped launch this wave when I was the
CTO and VP of Engineering at Wayport.   Stretching this convenience
function to a wide area deployment is the trap door in many a "wireless
broadband" biz plan.

> 3) "In urban areas WiMax does not make sense" for companies that have
> already spent billions laying copper.  Perhaps some competition would
> be nice.  As it stands, your ubiquitous Internet options are:

Note that in urban areas, deploying WiMax over unlicensed spectrum
won't work.  Nobody with 6 brain cells attempts to erect a (for-pay)
service over unlicensed spectrum in an uncontrolled interference
environment.   The companies who own (or can pay for) spectrum aren't
generally interested in deploying WiMax.

> Does someone see a market in here for a T1 with a big footprint?  I do.

So did vivato.  Should I re-tell that story?

>   And this is before anyone has unleashed any ingenuity on the problem.
> No one anticipated someone getting a few miles out of a WiFi base
> station and a Pringles can.  Or the competitions to get signal
> broadcast for 40 miles across the Mojave Desert.

Do you really want the technical analysis of why these things won't
scale to a service?


------ Forwarded Message
From: Jeff Porten <jeff () jeffporten com>
Date: Mon, 07 Feb 2005 15:01:30 -0500
To: Jim Thompson <jim () netgate com>
Cc: <dave () farber net>, Ip <ip () v2 listbox com>
Subject: Re: [IP] more on Sobering analysis of WiMax

I get the general impression that you and I are arguing apples and
oranges to a great extent.

On Feb 7, 2005, at 2:14 PM, Jim Thompson wrote:

> but of course, these types of "solution" are available today, and
> haven't made "wireless broadband" a reality.

I'm online more than 12 hours per day, and my laptop is frequently left
online 24/7.  With the exception of visits to client sites, I haven't
plugged an Ethernet cable for Internet access into my computer for
years.  I would call that reality.

Or are we talking about "ubiquitous wireless broadband with no digital
divide"?  I suggest that there is a large enough difference between the
two concepts that it's unfair to equate them so blithely.  In which
case, I agree that we're quite a ways off from ubiquity, but I note
that given how much has been done with the limited antennas of 2004, I
anticipate the larger ranges of 2005 and 2006 will create, if you will,
a network effect beyond what the pessimists expect.

> Repeat after me, "you'll never replace wires with radios".   There
> will always be more capacity in
> a wire than in a RF link.  Always.

Of course.  Forgive me for thinking this a somewhat condescending
comment, considering I work with wired/wireless network setup on a
daily basis.  But the question is not whether I have more bandwidth
over my IP-over-Firewire, but whether my wireless connection is *fast
enough* to handle whatever I might reasonably want to do over the
connection.  The vast majority of the market does not require 100 Mbit
wireless upstream, nor will they until some clever soul finds a useful
reason for them to push that much data.  But it can be awfully useful
at home.

> Yes, the convenience of not having to locate outlets make WiFi nice in
> coffee shops and airports.  I helped launch this wave when I was the
> CTO and VP of Engineering at Wayport.   Stretching this convenience
> function to a wide area deployment is the trap door in many a
> "wireless broadband" biz plan.

Far be it from me to argue with you about wireless business models,
then.  My own WiFi business plan got torched the day T-Mobile announced
a $2B Starbucks Hotspot investment; my partners and investors and I
rolled up our plans about 12 hours later.

That having been said, all of the current WISPs, including yours, built
out based on the technology that was available.  Newer technology leads
to new business models.  I switched my cell phone and two Internet
services over to T-Mobile largely on the strength of their early-2004
GPRS+Hotspot package.  The first company who offers me a T1 footprint
that covers all of DC will get a running start at my T-Mobile contract,
especially since by then I expect I'll have a WiFi cellphone to use
with my Vonage account.  Or maybe I'll have switched entirely to Skype.

> Note that in urban areas, deploying WiMax over unlicensed spectrum
> won't work.  Nobody with 6 brain cells attempts to erect a (for-pay)
> service over unlicensed spectrum in an uncontrolled interference
> environment.

Huh.  Interesting.  I guess all of those cordless phones, wireless
routers, and Bluetooth devices that I'm seeing are a mirage; they all
work over the same unlicensed spectrum, and if the home isn't an
uncontrolled interference environment then you don't know anyone who
owns a microwave.

If your point is that no one offers such services in fixed public areas
where they can't control QoS -- sure.  But you could have made much the
same argument about cell phones, which was solved by throwing a few
billion dollars at multiple overlapping towers.  And which, notably, to
this day has millions of customers who regularly just live with dead
zones.  If all people experiencing dead zones dumped their cells due to
QoS, then we'd all be walking around with extra quarters in our pockets
for the few pay phones that are left.

> The companies who own (or can pay for) spectrum aren't generally
> interested in deploying WiMax.

At least, not until some startup starts eating their lunch.

> So did vivato.  Should I re-tell that story?

If you like.

> Do you really want the technical analysis of why these things won't
> scale to a service?

No, I don't need someone to tell me why you shouldn't base a business
plan on the Pringles antenna.  My question is why you think this level
of innovation will suddenly stop with current technology.  If the
technology is available, the grassroots will innovate with it; if the
technology is not available, they'll do what they can to build it.

Put another way, if the large telecoms can't figure out how to make
something work profitably, we should not make the mistake of thinking
that that means it won't work at all.

Best,
Jeff


------ End of Forwarded Message


-------------------------------------
You are subscribed as srinivas () radiofiber com
To manage your subscription, go to
  http://v2.listbox.com/member/?listname=ip

Archives at: http://www.interesting-people.org/archives/interesting-people/



------ End of Forwarded Message

-------------------------------------
You are subscribed as lists-ip () insecure org
To manage your subscription, go to
  http://v2.listbox.com/member/?listname=ip

Archives at: http://www.interesting-people.org/archives/interesting-people/