Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

RE: [EFM] EFM / DOCSIS comparison


1. DOCSIS 1.1 is more about QoS, link data security and network management.
The MAC protocol
works on DOCSIS 1.0. FTTx will require the same functionality if it is to
deliver differentiated services in a cost effective manner.

2. Conserving b/w on 1 Gbps link will in the end be just as important as on
10 Mbps. I am not a great fan of wasting 20% of anything - and neither would
be the providers or their customers, as both would have to pay for the

3. The request/grant paradigm includes fixed size allocations without
explicit requests as a special case. In this sense the simple TMDA scheme
brings nothing new to the party.


-----Original Message-----
From: ramu [mailto:ramu_raskan@xxxxxxxxxxxxx]
Sent: Monday, July 16, 2001 4:17 PM
Subject: [EFM] EFM / DOCSIS comparison

(sorry if this is a duplicate. Earlier attempt to send did not appear to get
Harry, on this comment you made: "...those of us familiar with DOCSIS tend
to favor the dynamic request/grant protocol approach, while those that are
not, favor reinventing the wheel."

With all due respect to those that have worked (and are still working) on
getting certifiable DOCSIS 1.1 products, DOCSIS 1.1 is not a proven protocol
in real operation. It is still being "adjusted" in the labs. It may well
turn out just as Roy has said other attempts in the past have. Reinventing
DOCSIS is NOT what those trying to maintain a simple TDMA scheme seek.

It is also possible that those who have spent years buried in cable modems
are not "thinking outside the box" with respect to EFM and optical fiber.
There are some very dis-similar attributes. DOCSIS required a
conservationist view, as do all copper media. This is evident in the
emphasis on "waste" you made in the following post:

"However, wasting 20% on a Gigabit link represents 200 Mbps - compared to 2
Mbps on HFC. Hence the need to use intelligent b/w allocation on ePON

First, this is non-sequitur. Second, you are also assuming a 32-split node
(because it makes the comparison to DOCSIS look better?)
But, isn't it telling that EFM-over-fiber can WASTE more bandwidth across
16-32 users than DOCSIS has in total for 1000, and still outperform DOCSIS
by orders of magnitude?

It is not what we can save that matters, but what we can provide.

Get 250 color business cards for FREE!