Re: PMD discussion
Roy, Jay,
Please note that your responses are with respect to comments I made about
alternatives to the WAN PHY, not the SONET infrastructure. I don't view 1550 nm
solutions, WDM or Serial as being optimum solutions for the "WAN PHY"
application space. In that note (below) I said:
"Let's also not confuse the issue. The WAN PHY is not intended for use over the
SONET infrastructure. That's SONET's domain."
To elaborate a bit further, for 10 Gigabit Ethernet:
The LAN PHY is for use in the LAN;
SONET/SDH is the infrastructure in place for 10 GbE payload transport in the MAN
and WAN;
The WAN PHY is a specific PHY intended to bridge between the LAN and WAN using
the WIS as the bridge.
If the above statements are incorrect, please point out the flaws in them.
Best Regards,
Rich
--
jay.hoge@xxxxxxxxxxxxxx wrote:
>
> Roy,
>
> There is another issue to consider in long haul transmission at these
> frequencies; chromatic disbursion. Unless disbursion shifted fibre is used
> (which precludes the use of DWDM), 1310nm has much less chromatic
> disbursion, which translates into ISI. I'd still vote for 1550nm, because
> of the widespread availability of Telcom components, not to mention EDFA's,
> but a good case can be made for a high power direct modulated 1310 DFB link
> being cheaper for the same performance over SMF28 than an externally
> modulated 1550nm link.
>
> Jay
Roy Bynum wrote:
>
> Rich,
>
> The issue of 1500nm vs. 1300nm at 40km is link budget. For the longer distance, 1500nm solutions have lower attenuation in the same
> fiber as 1300nm solutions, which translates to a lower link budget requirement, which translates to lower launch power and better
> receiver sensitivity. Both the lower launch power and better receiver sensitivity translate into lower costs.
>
> I am sure that some of the optical component people will be able to put in specific details on this issue. In that I would like to
> see one of them generate a better spreadsheet that includes some of the relative cost points and a speculative time line of
> component cost changes.
>
> Thank you,
> Roy Bynum
>
> ----- Original Message -----
> From: "Rich Taborek" <rtaborek@xxxxxxxxxxxxx>
> To: "HSSG" <stds-802-3-hssg@xxxxxxxx>
> Sent: Monday, May 29, 2000 11:34 PM
> Subject: Re: PMD discussion
>
> >
> > Stuart,
> >
> > I never said that that locating the WIS at the end of a LAN PHY was the ONLY way
> > to connect 10 GbE to SONET. In fact, I used the word ALTERNATIVE in my
> > illustration. I believe its very important to point out all the alternatives in
> > a standards process, it helps us meet the PAR's 5 criteria.
> >
> > I'm just pointing out that the WAN PHY is not the ONLY way to provide SONET
> > compatibility. Alternatively, SONET compatibility can be achieved with the
> > simple and inexpensive LAN PHY and a WIS element used whenever it is needed.
> > It's the WIS that provides Ethernet-to-SONET layer 1 bridging, not the WAN PHY.
> > I'm all for standardizing the WIS and leaving implementations, including the WAN
> > PHY, out of the picture.
> >
> > That said. I'll turn your argument right around on you. It makes more sense that
> > way:
> >
> > Currently, ALL Ethernet links that go into the WAN, or SONET do not employ a WAN
> > PHY. Ethernet links to SONET are generally SONET, typically carrying
> > Packet-Over-SONET traffic. These links are typically much more expensive that
> > their Ethernet counterparts. This leads to a strong desire to cost reduce the
> > links.
> >
> > The WAN PHY proposal supports 10 GbE LAN connectivity to SONET, but by requiring
> > SONET-like (sorry, but a WAN PHY bears little if any resemblance to Ethernet).
> > Doesn't the WAN PHY sound like a specific implementation to solve a simple
> > problem.
> >
> > BTW, I'm having a difficult time with your argument with respect to the 1550 nm
> > PMD. I view 1550 nm in much the same way as the WAN PHY, a non optimal solution
> > for SONET compatibility.
> >
> > Let's also not confuse the issue. The WAN PHY is not intended for use over the
> > SONET infrastructure. That's SONET's domain.
> >
> > I heartily agree with your last statement: "There may be options to where the
> > WIS could be placed. However, the standard should not be such that it dictates
> > where."
> >
> > Are we perhaps in agreement?
> >
> > Best Regards,
> > Rich
> >
> > --
> > Stuart Robinson wrote:
> > >
> > > Hi Rich,
> > >
> > > I think that the confusion comes from the discussion of the standard versus
> > > one possible implementation. It seems like you are trying to have one
> > > implementation as the standard.
> > >
> > > If I follow your arguement, then Serial 1550nm optics may meet the distance
> > > objectives in the same way your LAN PHY + WIS may meet the WAN PHY
> > > objectives. However, these are not necessarily optimized for their
> > > respective applications.
> > >
> > > Your LAN PHY + WIS (with the WIS being in the transport equipment) is
> > > putting a long interface (ie SERIAL LAN PHY) between the PCS and WIS (stack
> > > diagram on Pg 4 of
> > > http://grouper.ieee.org/groups/802/3/ae/public/may00/bottorff_1_0500.pdf).
> > >
> > > One may argue that this meets the objectives but it is not necessarily
> > > optimized for applications that run over SONET infrastructure or "Optical
> > > Network" equipment (which are actually based on SONET).
> > >
> > > There may be options to where the WIS could be placed. However, the
> > > standard should not be such that it dictates where. Vendors may achieve
> > > Howard's Nirvana - LAN PHY, WIS, and SERDES in one device (p16 of
> > > frazier_1_0300.pdf link enclosed below).
> > >
> > > Regards,
> > >
> > > Stuart
> > >
> > > -----Original Message-----
> > > From: Rich Taborek [mailto:rtaborek@xxxxxxxxxxxxx]
> > > Sent: Monday, May 29, 2000 2:27 PM
> > > To: HSSG
> > > Subject: Re: PMD discussion
> > >
> > > Jay,
> > >
> > > I agree with you. However, I believe that we are loosing sight of the
> > > ultimate
> > > goal of the WAN PHY: Compatibility with the SONET/SDH infrastructure.
> > >
> > > I'd like to address a comment regarding our objectives that was made during
> > > the
> > > Ottawa meeting on several occasions: It was said that a LAN PHY with a WIS
> > > (for
> > > SONET/SDH compatibility) does not meet the HSSG objectives. I'm having
> > > trouble
> > > understanding why a solution which "exceeds" the objectives, and is highly
> > > likely to be lower in cost, is inferior to one which "meets" the objectives.
> > >
> > > The specific objective in question is as follows:
> > >
> > > Define two families of PHYs
> > > - A LAN PHY, operating at a data rate of 10.000 Gb/s
> > > - A WAN PHY, operating at a data rate compatible with the payload rate of
> > > OC-192c/SDH VC-4-64c
> > >
> > > It has been proposed, and there is general agreement (i.e. lack of any other
> > > proposal) that the WIS, a layer 1 (PHY) sublayer is used to encapsulate
> > > Ethernet
> > > packets, using 64B/66B for a PCS, for transport over SONET/SDH. The latest
> > > relevant proposals from Ottawa are:
> > >
> > > http://grouper.ieee.org/groups/802/3/ae/public/may00/bottorff_1_0500.pdf
> > > http://grouper.ieee.org/groups/802/3/ae/public/may00/nicholl_1_0500.pdf
> > > http://grouper.ieee.org/groups/802/3/ae/public/may00/walker_1_0500.pdf
> > >
> > > The location of the WIS in a WAN PHY can be illustrated as follows:
> > >
> > > +-----+ +-----+ +-----+ +-----+
> > > --->| |--->| PHY |------->| |--->|SONET|
> > > | MAC | | | medium | PHY | | \ |
> > > <---| |<---|(WIS)|<-------| |<---| SDH |
> > > +-----+ +-----+ +-----+ +-----+
> > > (rate
> > > controlled) |<--- WAN PHY ---->|
> > >
> > > Alternatively, the location of the WIS in a LAN PHY can be illustrated as
> > > follows:
> > >
> > > +-----+ +-----+ +-----+ +-----+
> > > --->| |--->| |------->| PHY |--->|SONET|
> > > | MAC | | PHY | medium | | | \ |
> > > <---| |<---| |<-------|(WIS)|<---| SDH |
> > > +-----+ +-----+ +-----+ +-----+
> > > (rate
> > > controlled) |<--- LAN PHY ---->|
> > >
> > > Note that the WIS function simply moves to the right (towards SONET/SDH) and
> > > the
> > > same PHY, a LAN "UniPHY" if you will, may be used to achieve full SONET
> > > compatibility. MAC/PHY rate control is not issue between the two methods. I
> > > understand how to implement it either way.
> > >
> > > Please point out the flaw(s) with the LAN UniPHY in supporting Ethernet over
> > > SONET. Is it that there is a unwritten requirement for a WAN PHY to support
> > > payloads other than Ethernet over SONET/SDH? I'd call any such requirements
> > > out
> > > of the scope of IEEE 802.3.
-------------------------------------------------------
Richard Taborek Sr. Phone: 408-845-6102
Chief Technology Officer Cell: 408-832-3957
nSerial Corporation Fax: 408-845-6114
2500-5 Augustine Dr. mailto:rtaborek@xxxxxxxxxxx
Santa Clara, CA 95054 http://www.nSerial.com