Hopes for Wireless Cities Fade as Internet Providers Pull Out

[David Farber <dave () farber net>]

Begin forwarded message:
From: dewayne () warpspeed com (Dewayne Hendricks)
Date: March 24, 2008 8:19:12 AM EDT
To: Dewayne-Net Technology List <xyzzy () warpspeed com>
Subject: [Dewayne-Net] Re: Hopes for Wireless Cities Fade as Internet  
Providers Pull Out

[Note:  This comment comes from friend Ken DiPietro.  DLH]

From: ken <ken () new-isp net>
Date: March 24, 2008 4:19:03 AM PDT
To: Dewayne Hendricks <dewayne () warpspeed com>
Subject: Hopes for Wireless Cities - The sky is not falling. (long)

Bob,

Thanks for your reply. As you can see, based on the passion I approach  
this subject, this is something I feel deeply about. My apologies, in  
advance, for the length of this reply and suggest that many people on  
this list might wish to skip past this post, should Dewayne elect to  
distribute this message to the list.


From my perspective, there are five distinct areas where we have  
sharp disagreement, including; business model, technology (as defined  
by the network architecture), the value of applications, what we  
should accept with respect to reliability, and lastly, philosophy. For  
clarity's sake, I will parse this reply by topic to present a far  
clearer understanding of my vision, as I believe that the overall  
difference between us is one of what we believe the future should look  
like.

There is one point that I would like to make abundantly clear. There  
is a need for more viable options for these critical services than we  
have to choose from currently. The corporations now operating in this  
space place significantly higher value on profit over public safety,  
which is to be expected. However, this is something that we, as a  
society, should have learned is not in our best interest based on the  
lessons provided by 9/11 and Katrina.


The Business model -

We certainly agree that most of the attempts at building these  
MuniWiFi networks to date been based on horrific business models. One  
might want to take into account that this experiment is less than five  
years old and that when an experimental network fails it is to be  
expected, or rather embraced, as it is through efforts like this that  
we all learn.

With that said, I believe there are several business models that will  
flourish and remain sustainable, but every one of them is dependent on  
the underlying technology chosen. WiFi, especially Ad Hoc networking,  
is exactly the wrong technology to use in deploying the type of  
networks I envision.

Let me offer this premise, the communications industry thrives based  
on one crucial fact - businesses will pay more for services, as long  
as those services are both cost effective and 100% reliable.

If we are willing to accept that, we can then move on to figuring out  
how we can lower the cost for communications for society as a whole -  
something we both share a belief in.

As a specific example, gigabit connectivity in the major metropolitan  
US markets is priced somewhere around $10K per month. At today's  
equipment prices, we would expect to see roughly a three month ROI to  
deliver this service to this customer. Based on historical evidence we  
can expect an 80% price drop in this equipment once it hits mass  
production. This cost reduction should then translate into a  
substantial decrease in fees to this customer. Please note, this one  
customer has also very willingly subsidized a one gigabit loop in the  
network while lowering their cost for service. This is the definition  
of a win/win business arrangement.

Technology -

In a time where single digit Mbps connectivity places our country at  
the bottom of the industrialized world, you are advocating that we  
deploy a
*shared* single digit Mbps network. To add insult to injury, these Ad  
Hoc network operates in half duplex mode in a world where applications  
like VoIP and Video Conferencing demand full duplex, symmetrical, and  
low latency connections. Further, when used in a "Mesh" architecture  
we have a series of half duplex radios "relaying" full duplex  
applications, across an unknown number of hops, with each hopping  
device (adding latency at every hop) while including in their own  
traffic load. My experience tells me that this type of hardware is  
unsuitable for this kind of work and that in times of increased need,  
these networks fold under the onslaught of relatively few VoIP calls  
alone. Add to that the reality that these networks must all use the  
same channel, one can see that the busier the network becomes the more  
self-interference they create.

Instead, I see a very different future. As I type this to you, all of  
the pieces are available to create the network I am about to describe,  
even though, admittedly, the cost for some of the components are quite  
high currently.

At the core of this proposed network is fiber - and in most urban  
areas of this country, if not all, there is an adequate supply or  
whatever is missing can be built out, as necessary. This fiber is now  
fed into ultra high capacity wireless links such as the kind  
manufactured by Gigabeam, which are capable of delivering speeds in  
excess of one gigbit full duplex. These ultra high frequency radios  
have a relatively short range but are perfect for creating the first  
layer off of our fiber core. This "EBand" technology also has the  
ability to collocate "near limitless" links as self-interference is  
almost non-existent. Based on our ability to carry multi-gigabit  
connectivity across this layer, we can now provide customers who need  
this service a product that they will gladly pay for.

The next level down offers many choices in both licensed and license  
exempt hardware which are capable of delivering throughput in the  
multiple hundreds of megabits up to in excess of one gigabit. In any  
given urban area these radios could be deployed to cover the next  
layer of this network, allowing for SLA capable service to be deployed  
for businesses that are interested in paying. This layer, as well as  
the EBand layer previously described, serves a dual purpose allowing  
us to connect the edge network into the fabric.

At the edge, we new deploy shared connectivity of 100Mbps, or greater,  
for fixed/mobile communications. Given the current adoption of WiFi  
technology we need to accept that WiFi is the common connection  
vehicle and that admittefly while this is not conducive to delivering  
speeds in excess of 50Mbps, that day is rapidly approaching with the  
rise of the draft n standard. Until that time, multiple access points  
will need to segregate the slower (conventional) WiFi users from the N  
users so as to provide faster speeds for those who are able to take  
advantage of them and a slower class of service (for VoIP and other  
low bandwidth applications like meter reading) to share.

Returning back to the business model (as these sections are  
intertwined) one can see how a potential for a marketable service  
offering springs to life when we realize that we are not talking about  
a miserable, DSL-like service but rather a service that bests the  
fastest commercially available service, assuming we give Fios that  
award.

Please note - As this discussion has completely omitted any mention of  
bringing connectivity into rural areas I have not addressed this  
issue. This is not to say that I do not believe this is a critical  
problem and perhaps we can revisit this pressing issue in another  
thread, if the list is interested.

Applications -

Let's narrow this view to one specific application of this technology  
as an example so as to differentiate our visions.

Imagine a city where every first responder is equipped with ubiquitous  
full, multimedia capable, mobile connectivity. This now provides for a  
real level of situational awareness to be implemented and in the case  
of paramedics, telemetrics could be transmitted, not just to the  
emergency room nurse but to specialists for consultation during the  
ambulance ride or Medivac trip to the hospital. In the case of a  
stroke this could make all the difference between a recovery and a  
vegetable.

Moving out from that scenario, 802.11p offers the promise of real time  
communications between the emergency vehicle and other drivers for  
increased safety. This makes it possible to warn non-emergency  
vehicles of an approaching emergency vehicle and/or changing the  
traffic signal to allow for an uninterrupted travel to an emergency,  
when every second counts. Even better, in cases of traffic jams or  
accident the ambulance could be rerouted around the problem, in real  
time, allowing it to reach the hospital in as little time as possible.  
This is what you alluded to with your comment regarding real time  
traffic monitoring.

Is this a service that people will pay for? No, not directly, but the  
huge sums of money now being spent to provide emergency communications  
could be shifted to help offset the cost of building these networks  
with a net gain in usability and functionality. This is only one  
example of a service that improves the quality of life and I think we  
both agree that money isn't everything.

Currently, I am tracking close to two dozen applications that either  
add value to a well designed network (read generate revenue) or  
improve the quality of life for residents of any area this technology  
is deployed in
- with more applications being added in on a regular basis.

Reliability -

Briefly, we have gone from a country that took our nearly 100%  
reliable communications network for granted, one that was once the  
communications network the rest of the world measured against, to one  
that puts up with dropped calls, miserable sound quality (like anyone  
thought it could actually get worse) and areas of no coverage.  
Incredibly, we now have Verizon using this stupidity in their  
advertising campaigns and quite frankly I believe it is an admission  
of failure that they have to ask if I can hear them now.

I, for one, am not interested in "hearing" how we can continue this  
mindset. I think we have done an incredible job of degrading what we  
once had right.

Philosophy -

Here is where I am afraid we will see the greatest difference in our  
positions - one that I am at a loss to understand how an obviously  
intelligent person like yourself could possibly try to defend.

As I understand your position, you are proposing that we should  
"share" access points connected to residential connectivity to provide  
our poor with, what can only be considered, substandard connectivity.  
I find that to be nothing short of appalling.

In a world where the ability to get a great education, versus an  
adequate one, can be measured in the difference between a child having  
a rich, multimedia online experience, unimpeded by connection speed  
may literally be the difference between success and failure. What you  
are proposing, to my admittedly limited way of thinking, is that the  
wealthy get FTTH and the poor eat virtual cake. I am sorry, I find  
that nothing short of reprehensible.

I grew up during a time when building something of quality was what a  
person took pride in, a time when building something substandard was  
frowned upon, labeled as "cheap" and considered unacceptable, a time  
when an artisan returned home from a day's work feeling satisfied  
based on knowing that an exemplary job had been done.

Nowhere in your position can I find that set of values. Perhaps I am  
outdated and you are right in your views. If so, I will continue to  
advocate doing the wrong thing.

I look forward to your reply,

Ken


On Sat, 2008-03-22 at 23:55 -0700, Dewayne Hendricks wrote:
> [Note:  This comment comes from friend Bob Frankston.  DLH]
>
> From: "Bob Frankston" <Bob19-0501 () bobf frankston com>
> Date: March 22, 2008 2:22:44 PM PDT
> To: "'Dewayne Hendricks'" <dewayne () warpspeed com>
> Cc: "'ken'" <ken () new-isp net>
> Subject: RE: [Dewayne-Net] Re: Hopes for Wireless Cities Fade as
> Internet Providers Pull Out
>
> I’m respond to Ken and you can forward it to the list.
>
> The crux of the disagreement comes at the end of the comments with the
> focus on reliability and sigmas. The problem with sigma reliability as
> with QoS is that the measures are arbitrary and more about perfecting
> what we have than creating opportunity for much more.
>
> This is an issue I address in http://www.frankston.com/?name=WiFiEdge
> because it’s at the heart of the difference between the Telco idea of
> providing reliable services and the Internet focus on creating
> opportunity. The End-To-End argument (http://www.reed.com/dpr/docs/Papers/EndtoEnd.html
> ) explains how you can create solutions without depend on the middle
> for more than best efforts. I also explain more of how this plays out
> in http://www.frankston.com/?name=InternetDynamic and other writings.
> There’s also http://www.frankston.com/?name=Railroads in which I
> distinguish the railroad/service model from the Internet which is more
> akin to driving.
>
> I realize that much of what I’m saying is counter-intuitive and that
> it’s hard to convey decades of first-hand experience in watching these
> dynamics play it.
>
> I don’t want to go over each example but to use one – the Wi-Fi VoIP
> phone. If you want to guarantee it works you need to duplicate much of
> the high priced cellular network and you have to get people to pay for
> that and they will if they are dependent upon it. That is akin to have
> required voice work in 1985. Instead we found out what worked and
> built upon it.
>
> One day one ad-hoc wireless coverage will be far less than the current
> cellular coverage but it will also have an incremental cost of close
> to zero but you set in motion a dynamic that can deliver increasing
> capacity at a marginal increase in cost of close to zero. And it’s all
> about the dynamic and not the accidental properties of today’s
> technologies and deployments.



-------------------------------------------
Archives: http://www.listbox.com/member/archive/247/=now
RSS Feed: http://www.listbox.com/member/archive/rss/247/
Powered by Listbox: http://www.listbox.com