home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.dcom.cell-relay
- Path: sparky!uunet!usc!zaphod.mps.ohio-state.edu!sol.ctr.columbia.edu!usenet.ucs.indiana.edu!robelr@mythos.ucs.indiana.edu
- From: "Allen Robel" <robelr@mythos.ucs.indiana.edu>
- Subject: Comment made at Next Generation Networks conference
- Message-ID: <By1CzL.EL4@usenet.ucs.indiana.edu>
- Sender: <robelr@mythos.ucs.indiana.edu>
- Organization: Indiana University
- Date: Fri, 20 Nov 1992 22:20:00 GMT
- Lines: 28
-
- At the Next Generation Networks conference John
- McQuillan made the comment that 100Mb/s FDDI delivers
- more bandwidth than 100Mb/s ATM. Now I know the reasoning
- behind this (proportionally, overhead is less for FDDI
- variable-length frames depending on average frame
- size) but has anyone done any studies with real-world
- networks to determine how much real throughput FDDI
- delivers to a given individual station given that its a
- shared medium? How would one calculate the true
- throughput for any given station (raw theoretical
- throughput, not application layer throughput)?
-
-
- I would imagine possible variables that would enter a
- station throughput calculation might be the following:
- number of stations, token rotation time, token holding
- timer but how would one combine these (or others) into a
- FDDI station throughput formula? My motive here is to
- compare ATM's inefficiencies (cell overhead, SAR/AAL
- overhead???) with FDDI's inefficiencies (shared
- medium and what little frame overhead there is) as
- a reality check for McQuillan's statement.
-
-
-
- Thanks in advance!
-
- allen
-