Cable Tech

Ethernet Differences Revealed in Poll

Big regional differences in the target markets of carriers deploying Ethernet access technologies have emerged in Light Reading's September Research Poll.

In North America, Ethernet-to-the-home is considered a dead duck by respondents so far. It's got zero votes in the poll, while 65 percent of respondents expect Ethernet to have the biggest impact in services to small-to-medium-sized businesses.

In contrast, Ethernet-to-the-home gets the biggest vote -- 53 percent -- in the Asia/Pacific, with the small-to-medium businesses market garnering a mere 26 percent.

Western Europe occupies the middle ground. Ethernet-to-the-home gets 25 percent of the vote, while SMBs get 40 percent.

Other results of the poll include:

  • Ethernet will become more commonplace than ATM in copper access lines supporting broadband services in 2007, according to the largest proportion of respondents -- 35 percent.

  • Ethernet will take longer to become the predominant method of terminating telecom services over fiber access lines. The largest proportion of respondents think this won't happen until 2008 or later.

  • The primary incentive for telcos to roll out Ethernet access to businesses is being able to bundle multiple services on a single access line, and being able to use low-cost equipment (each got 30 percent of the vote).

  • The primary incentive for telcos to roll out Ethernet access to homes is bundling multiple services on a single access line (25 percent of the vote), followed by offering video-based services (20 percent).

  • The primary obstacle to Ethernet access to businesses is the threat of cannibalizing existing services (45 percent of the vote).

  • The primary obstacle to Ethernet-to-the-home is lack of user demand (25 percent). (FastWeb SpA in Italy would probably disagree -- see TV Over DSL Over Italy .)
To take the poll yourself and see the latest results in detail, please click on this link.

— Peter Heywood, Founding Editor, Light Reading
PO 12/4/2012 | 11:22:59 PM
re: Ethernet Differences Revealed in Poll Once again the results of a survey are made more difficult than necessary to interpret because the terminology may mean different things to different people depending on their level of awareness.

While I doubt very much that anyone was confusing CSMA/CD with MAC-layer PDUs, the presence or absence of Ethernet in today's Access network seems to be a matter of opinion as much as fact.

A few years ago a well-respected data networking expert was very disappointed to learn that his access service was implemented using ATM. As far as he was concerned, he plugged an Ethernet cable into his (DSL) modem and got IP connectivity. But is Ethernet present in this access network? Some say yes, many say no.

As near as I can guess, the survey was trying to ask about the importance of a MAC PDU-optimized layer 2 access link. Surely it wasn't intending to suggest that perhaps we should be installing ATM NICs because nobody voted for Ethernet as being important in North American homes.

And I'm also pretty sure the survey wasn't trying to only motivate the use of Ethernet as a layer one technology. Whatever that means to you.

Then, of course, without defining the context for Ethernet within the access Network, there is room to interpret the questions as asking about the importance of Ethernet Networking (layer 2 switching, as opposed to IP routing which is a layer 3 activity) within a neighbourhood. But if Ethernet is only being referred to as a point-to-point access link technology, then what are the survey questions really asking?

While the survey questions appear meaningful, the more I reflect on them the harder it is to draw useful conclusions.
Ramu3 12/4/2012 | 11:22:52 PM
re: Ethernet Differences Revealed in Poll LR: I'd love to see a similar poll about fiber to the home barriers, threats, perceptions, geographical differences, business model possibilities, etc.

What would LR think about setting up a separate thread to have readers come up with suggested questions for the poll?

Maybe it could be more general, such as: how much broadband is enough, why, where, how much will it cost, who will pay, what technology, what are the scalability requirements, QoS, bandwidth symmetry, operating structure of the network (i.e. vertically-integrated versus structurally separated and why), what explains the vastly different take rates across geography, impacts of competition at various layers (ducts, fiber, local access network equipment, services), perceptions of municipal and utility companies deploying and/or operating such networks, and so forth.

Then, maybe a specific section on fiber, wireless, copper, and hybrids.

I think you'd have a very marketable report with this flow. It would be unique, relative to all the other copy-cat reports about broadband that all say the same thing. Plus you'd get a very diverse set of perspectives from readers that would let you boil the issues down to some fundamental questions.

If people could chime in here and let the LR staff know they'd be interested in participating (so LR has a feel for the level of interest), please do.
telebud 12/4/2012 | 11:22:50 PM
re: Ethernet Differences Revealed in Poll "But if Ethernet is only being referred to as a point-to-point access link technology, then what are the survey questions really asking?"
The survey does not bring in the CONSTRAINTS
of an ETHERNET NETWORK repeaters needed every 185m and Bridges for congestion.
connection from a single user the signal still has to be sent accross the access wire.
If you have a neighborhood with 500 houses
say several streets connecting to a
single access provider box how much hardware
would that take to implement?
The last Ethernet Mile could be a BIG PROBLEM
to manage.
The survey mentioned ETHERNET IMPACT for business Maybe using a SMALL METRO ETHERNET LAN like in a

PO 12/4/2012 | 11:22:49 PM
re: Ethernet Differences Revealed in Poll Telebud: "...ETHERNET NETWORK repeaters needed every 185m and Bridges for congestion"

PO: Sounds like you're referring to 10BaseT and 100BaseTX. 100BaseFX would imply different constraints.

I'm not convinced the survey was defining the physical medium (balanced pairs, point-to-point fiber, PON, coax, wireless). It's pretty rare these days for people to be referring to one specific physical layer when talking about Ethernet. But if that's what it means to you, fine.

Anyone have data on crosstalk for a bundle of, say, 100 pairs carrying 100BaseT?

The problem remains that the more we look at the survey the less conclusive the questions appear.
rjmcmahon 12/4/2012 | 11:22:49 PM
re: Ethernet Differences Revealed in Poll Along similar lines, and in LR traditions, it would be interesting to have a top 10 list of cities, counties, and regions which revealed the leaders in the deployment of modern communication infrastructures. Tie that data to things like a regions GDP and education levels and see if communcations abundance really helps the advancement of the human species.
Sign In