home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.dcom.cell-relay
- Path: sparky!uunet!munnari.oz.au!titan!titan!lampard
- From: lampard@titan.trl.OZ.AU (Greg Lampard)
- Subject: Re: WDM and ATM
- Message-ID: <1992Dec16.230531.27476@trl.oz.au>
- Sender: root@trl.oz.au (System PRIVILEGED Account)
- Organization: Telecom Research Labs, Melbourne, Australia
- References: <1992Dec14.221325.2405@trl.oz.au> <1992Dec16.053840.10120@atri.curtin.edu.au>
- Date: Wed, 16 Dec 1992 23:05:31 GMT
- Lines: 54
-
- clarke@atri.curtin.edu.au (Nathan Clarke) writes:
-
- >In article 2405@trl.oz.au, lampard@titan.trl.OZ.AU (Greg Lampard) writes:
- >> mark@alsvid.nr.une.edu.au (Mark Garrett Internet: mark@arvak.une.edu.au Phone: +61 66 20 3859) writes:
- >>
- >> >From article <1992Nov30.215429.29379@sics.se>, by craig@sics.se (Craig Partridge):
- >> >>>From: peterd@pjd.dev.cdx.mot.com (Peter Desnoyers)
- >> >>> Secondly, people I've
- >> >>> talked to who work with WDM believe that ATM is fundamentally
- >> >>> incompatible with the more promising WDM switching mechanisms, due to
- >>
- >> > I've come across ATM, but what is WDM? What are the specs for WDM?
- >> >from the jist of the posting I guess WDM is some form of fast packer switching
- >> >with "Wider" packs ???
- >>
- >> > Mark
- >>
- >> From the limited information above, I'd say that WDM refers to
- >> optical wavelength division multiplexing. That is, using different
- >> wavelengths of light over optical fibre to provide medium access to
- >> multiple users. It's analogous to the way radio spectrum is currently
- >> used, and obviously is confined to all-optical networks.
- >>
- >> Greg.
-
- >This is true: WDM = Wavelength Division Multiplexing. But I'm not sure
- >why "ATM is fundamentally incompatible with the more promising WDM
- >switching mechanisms". WDM is an OSI layer 1 (physical layer)
- >multiplexing mechanism whereas ATM is at layer 2. WDM (I thought) is
- >used merely to mux multiple channels (bit streams) onto one optic
- >fibre. The channels themselves may/mayn't have undergone ATM
- >multiplexing; there should be no dependency.
-
- >Nathan Clarke. Australian Telecommunications
- > Research Laboratory (ATRI) .-_|\
-
- I think the incompatibility being referred to is that a "WDM switching
- mechanism" uses the wavelength of light being transmitted to route information
- through the network, as opposed to ATM which uses headers. That is, in
- principal in an all optical network you don't have to have switches
- reading header information, they just have to be configured so that
- light of a given wavelength gets passed to the appropriate output. Such
- a switch is transparent to the data being transmitted, so it will work
- with any protocol and data rate. So, for example,
- you can send ATM cells over a WDM network, but the WDM switches wouldn't use
- the cell-headers for routing, you'd need another signalling system for
- that.
-
- That being said, there are a lot of problems to be overcome before networks
- of this type will be a commercial proposition. The most serious one
- being that optical components, particularly wavelength-tunable ones, are
- not cheap.
-
- Greg.
-