Re: SJTP: Minutes from 24-April meeting
All,
Attached is the comparison information I've pulled out of the
3 proposals (Ewen, Thatcher & Thaler). Please review and be
prepared to discuss tomorrow. Also, note the motion I added to
the top of the attachment. We're supposed to release our new
proposal for a 2 week circulation period before the May meeting.
This only gives us 2 more conference calls. Ideally, I'd like
to come out of this meeting with an action for someone to put
together a proposal that we can review at the next meeting.
After that next meeting, the proposal is released for circulation.
Thanks,
Ben
--
-----------------------------------------
Benjamin Brown
AMCC
2 Commerce Park West
Suite 104
Bedford NH 03110
603-641-9837 - Work
603-491-0296 - Cell
603-626-7455 - Fax
603-798-4115 - Home Office
bbrown@xxxxxxxx
-----------------------------------------
30-April-2001
Motion # 2 General Session Motion
(from unapproved minutes of March Plenary task force meeting)
Description: Move to create an ad-hoc to bring to the May
meeting (with circulation 2 weeks before the meeting) a
complete set of changes to the draft and justification with
consideration of the existing scrambler, ewen_1_0301,
thatcher_1_0301 and thaler_1_0301 presentations.
Motion Type: Technical > 75% required
Moved By: B. Reysen
Seconded By: P. Thaler
Results: All Attendees Acclimation 100%
Time: 11:32am 03/13/01
P/F: Passed
This motion means we need to have a proposal available to
the task force within the next 2 weeks for a 2 week circulation
period. We've got some work to do...
Common Themes:
* Use the existing 1+x39+x58 scrambler
* Seed the scrambler with a starting value
* Operate the scrambler for a finite period of time
* Use the 66-bit framing with a fixed sync header value
* Count errors at the receiver
Differences:
* Fixed vs configurable data input
- One suggestion is to use the local fault encoding which
works well if one side is in test mode while the other
side is in normal mode. Is this really a viable concern?
* Fixed vs configurable scrambler seeds
- I think we've already agreed to allow for configurable
seeds in order to have flexible patterns to satisfy the
"Pattern Content" feature from our feature list.
* The number of seeds/pseudoseeds
* Fixed vs configurable pattern durations
"Don't Cares"
* Count the errors by block or bit.