Interesting People mailing list archives

more on Is broadband set to make power lines sing? - alreadydoes in the UK


From: Dave Farber <dave () farber net>
Date: Wed, 25 Feb 2004 09:50:39 -0400


-----Original Message-----
From: Rahul Tongia <tongia () andrew cmu edu>
Date: Wed, 25 Feb 2004 08:47:32 
To:dave () farber net
Subject: Re: [IP] more on Is broadband set to make power lines sing? - already
 does in the UK

Dave,

A few observations [if readers wish, I presented an analysis of PLC at TPRC 
2003:
http://intel.si.umich.edu/tprc/papers/2003/246/Tongia-PLC.pdf
]

1) This boils down to not technology, but the business case and execution. 
Most consumers (and being a shared medium like cable implies largely 
consumer usage) really care about price.  Electricity's 99.98% reliability 
is not what they care about.
2) The US is particularly at a disadvantage when it comes to PLC (or BPL as 
some call it) since the number of homes per distribution transformers is 
typically 4-7, unlike Europe or Asia where it can be 30-50 times higher. 
The use of PLC as an uplinking backhaul technology with wireless last few 
hundred feet implies a limit of 10s of Mbps (and eventually, 100 Mbps) for 
the entire feeder line coming out of the substation, which is okay for the 
hundreds of potential users, but not earth-shattering. Deep fiber, like in 
Japan and Korea, may be more important if we aim to do the triple play, 
which certainly requires more bandwidth than the competition for "basic 
broadband" (i.e., <1 Mbps).
3) I find it misleading that that the FCC (and proponents) tout PLC as the 
alternative for rural areas.  PLC becomes much more expensive when we have 
to do long runs.  Sure it can be done.  But at a price, just like other 
copper-based solutions.  WiMax or the equivalent may be a better solution.

Rahul
 tongia () cmu edu


--On Wednesday, February 25, 2004 6:30 AM -0500 Dave Farber 
<dave () farber net> wrote:


Delivered-To: dfarber+ () ux13 sp cs cmu edu
Date: Wed, 25 Feb 2004 09:06:12 +0100
From: Andrew Bale <andrewbale () hotmail com>
Subject: Re: [IP] Is broadband set to make power lines sing? - already
does in   the UK
X-Originating-IP: [213.69.211.246]
X-Sender: andrewbale () hotmail com
To: dave () farber net

There have always been three major problems for powerline:
1) can you get data reliably over the (very noisy) power cables themselves
2) can you do it without using so much transmit power that it interferes
with every device in the neighbourhood
3) can you do it economically

(1) has (in most deployment scenarios) been solved by the use of advanced
signal processing techniques (for example home powerline LANs now work
very well - though have been completely overtaken by the even greater
convenience of in-home wireless)
(2) is largely solved - better receive signal processing reduces the need
for power in the transmitter
(3) remains - largely due to the architecture with which power is
distributed

Power systems rely on step-down transformers located close to the
consumer - each tending to serve a relatively small number of homes.
It's practically impossible to get a data signal through a transformer(!)
which means that data has to be passed from the core network to a
head-end device on the far side of the transformer.  The hardware and
deployment costs of this head-end device has to be shared amongst the
users of the system.  Since the transformers are generally only supplying
a small user base the economics demand a VERY high take-up - something
that experience shows is highly unlikely, unless you are in an area of
zero competition.  With all the competing options (cable, DSL, satellite)
that is unlikely.  While the UK trials below are clearly working, that
does not imply that they are demonstrating a good commercial model.
Power companies (especially in the UK and Germany) have been trying this
for years, largely in attempts to see whether they can persuade a high
enough density of customers to adopt the service to solve this economic
problem.

Interestingly - the service described in a previous [IP] mail used
wireless for the last leg to the customer - which (I think) implies that
they are inverting the more usual powerline scenario, where fiber is used
to the distribution box and powerline gets to the customer (as in the UK
trial below).  In the wireless scheme I think they are proposing to use
powerline over the main distribution network to replace fiber, and then
using wireless instead of powerline in the local loop.  Interesting, but
hardly earth-shattering.

Andrew Bale

----- Original Message -----
From: "Dave Farber" <dave () farber net>
To: <ip () v2 listbox com>
Sent: Tuesday, February 24, 2004 7:13 PM
Subject: [IP] Is broadband set to make power lines sing? - already does in
the UK



Delivered-To: dfarber+ () ux13 sp cs cmu edu
Date: Tue, 24 Feb 2004 18:05:18 +0000
From: "Charles Arthur, The Independent" <carthur () independent co uk>
Subject: Re: [IP] Is broadband set to make power lines sing? - already
does in
   the UK
X-Sender: Charles_main@localhost
To: dave () farber net

Hi ...

At 11:51 am -0500 on 24/2/04, you wrote:

  ><x-flowed>
  >
  >Is broadband set to make power lines sing?

It already does, in the UK. From an article in The Independent on February
4 (now behind a paywall on the site, so here's an extract):

".....
Scottish and Southern Energy have chosen Winchester and Stonehaven (near
Aberdeen) for full-scale commercial trials following successful pilots in
Crieff and Campbeltown. The company is expecting up to 1,000 users in
Winchester alone. Professor Cochrane's [former head of BT Research] views
are clear: "The laws of physics have not changed and the telephone line
well outperforms them. What these guys have and do promise, and what they
deliver, are worlds apart."
So who's right? The answer, of course, is to spend some time in Winchester
finding out.  Among the first users - just 33 of them in early December -
are Liz Galfskiy and her son, Adam. Mrs Galfskiy relies on her computer
for an Open University course. A fixed price dial-up AOL connection was
essential, especially as Adam is a keen user too. Several broadband offers
had left her unmoved until she heard about the Southern Electric trial.
In mid-November, engineers installed a booster box adjacent to her
electricity meter and provided a powerline broadband modem. For most
properties nearer a substation, only the modem is needed. Setting up the
broadband connection is trivial; plug in the modem (£50) and connect it
with an Ethernet cable to your PC. The service (£29.99 monthly) includes
unlimited e-mail addresses and web space too.
But is it strange to have your broadband connection through a 13-Amp
socket? Mrs Galfskiy doesn't think so. Thanks to her
technically-knowledgeable son, who installed a router, several PCs share
the single connection. Broadband allows them to watch BBC news items
online and listen to "crystal clear" radio: "It's excellent. I'd throughly
recommend it," she says.
A test with the speed checker on adslguide.org.uk confirmed an impressive
980 Kbps download - 20 times dial-up speeds. The upload result was a poor
478 Kbps given the advertised "up to" 1 Mbps for both directions. Adam
claims he's seen 1.7 Mbps connections.
But what happens when you turn on a hairdryer, run the washing machine or
vacuum the carpet?  "We haven't noticed any effects on the connection,"
responds Mrs Galfskiy.  However, choose the electrical socket with care
as, in her experience, powerline broadband dislikes multiple adaptors. .."


Not everything gets implemented first, or even best, in the US, you know.



          best
          Charles
--
   -------------------------------------------------------------------
The Independent newspaper on the Web: http://www.independent.co.uk/
          It's even better on paper

______________________________________________________________________
This email has been scanned by the MessageLabs Email Security System.
For more information please visit http://www.messagelabs.com/email
______________________________________________________________________

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

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

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

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




************************************************************************
Rahul Tongia, Ph.D.
Systems Scientist

School of Computer Science (ISRI) / Dept. of Engineering & Public Policy
Carnegie Mellon University
Pittsburgh, PA 15213 USA
tel: 412-268-5619
fax: 412-268-2338
email: tongia () cmu edu



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

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

Current thread: