RE: [EFM] EFM Requirements
Hi,
I read the charts from the Elastic presentation as well and I find it hard to analyze the results. Without understanding the assumptions used in the simulations it is not possible to compare them to the results from "standard" tools. I prefer using capacity calculations and not technology specific simulations to analyze the maximum reach/range that is possible in a specific environment. In the DSL standard groups this methodology was adopted. Once the capacity calculation parameters were fixed, every technolgy "camp" agreed with each other on the same theoretical limit. Both DMT and QAM companies agreed quickly to the limits of the cable plant. Now it is up to each to prove who gets closer to the theoretical limit. Any technology such as VDSL or Etherloop will not exceed this physical limit and it is a matter of the technology to define the implementation loss with respect to the maximal limit.
The agreed gap of 12 dB with respect to Shanon capacity was agreed to by operators, DMT companies and QAM companies. This assumes a coding gain of 3.8dB and a noise margin of 6dB. All the companies then assumed 20 additional systems in the cable binder. All the systems did not create NEXT to each other (FDD or TDD) but did create self FEXT. I do not know how the presentation translated the bursty nature of Etherloop. If this is explained and the simulation parameters will be provided then it is simple to model it in a capacity calculation that will be technology independant.
The details of the technology are not important to verify the results, only the capacity calculation assumptions.
Steven
-----Original Message-----
From: Fletcher E Kittredge [mailto:fkittred@xxxxxxx]
Sent: Wednesday, August 22, 2001 8:51 PM
To: Steven.Haas@xxxxxxxxxxxx
Cc: pstanley@elastic.com; stds-802-3-efm@ieee.org
Subject: Re: [EFM] EFM Requirements
On Wed, 22 Aug 2001 16:33:43 +0200 Steven.Haas@xxxxxxxxxxxx wrote: >
>This indicates the importance of understanding the tradeoff between
>distrance and rate. If the EFM wants to create a copper standard that
>support video streaming, it must reduce the range expectations
>accordingly. Rates in the area of 20 Mbps will not exceed 4500 feet.
I may be reading the chart wrong, but it appears that Elastic Networks
"Etherloop2", a 100mb/sec over copper protocol, claims to do better
than 20mb/sec at 4,500 feet. Do I have this wrong?
regards,
fletcher