home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.dcom.cell-relay
- Path: sparky!uunet!mcsun!sun4nl!research.ptt.nl!walvdrk
- From: walvdrk@research.ptt.nl (KEES VAN DER WAL)
- Subject: Re: Congestion Avoidance
- Message-ID: <1992Jul27.210256.1@research.ptt.nl>
- Sender: usenet@spider.research.ptt.nl (USEnet News)
- Nntp-Posting-Host: dnlts0
- Organization: PTT Research, The Netherlands
- References: <1992Jul18.230550.3046@sics.se> <1992Jul19.130005.1@research.ptt.nl> <1992Jul25.092823.12189@e2big.mko.dec.com> <k3jm90-.nagle@netcom.com> <miw.712192253@cc.uq.oz.au>
- Date: Mon, 27 Jul 1992 20:02:56 GMT
- Lines: 42
-
- In article <miw.712192253@cc.uq.oz.au>, miw@cc.uq.oz.au (Mark I. Williams)
- writes:
-
- > You're right. Congestion avoidance is still the big bugbear for ATM.
- > There is still (or wasn't 8 months ago) a generally accepted algorithm
- > for bandwidth allocation that performed better than peak-rate
- > multiplexing, although everybody is sure that some degree of statistical
- > multiplexing will be possible.
- >
- > Of course, that situation is where we are now. With the current
- > circuit-switched network we are using peak-rate multiplexing by
- > definition.
-
- But it multiplexes! I don't think the term "statistical multiplexing"
- (or even -sm- gain) is saying very much. Present day telephone networks are
- also multiplexed; unscheduled so "statistically".
-
- If operator X would run an ATM network with plain (and safe) peak rate
- allocation and if we could make very fast signalling procedures to renegotiate
- the capacity on an already established connection, what would then be the
- difference with network Y that would allow "statistical multiplexing" in
- the ATM-sense?
-
- It's like trading in congestion probability (in Y only) against the probability
- that the request for an increase in capacity is turned down (in X). I guess
- that X is to be judged better than Y as during congestion a large number of
- connections are going to suffer, as opposed to maybe a few in case X.
-
- Regards, <kees>
-
-
-
- Kees van der Wal e-mail: J.C.vanderWal@research.ptt.nl
- ----------------------------------------------------------------------------
- PTT Research Neher Laboratories Room: E130
- Sint Paulusstraat 4 Fax: +31 70 3326477
- 2264 XZ Leidschendam The Netherlands Phone: +31 70 3326295
- ----------------------------------------------------------------------------
-
- Nullum magnum ingenium sine mixtura dementiae fuit. -- Seneca
-
- Just guessing: "No bright ideas without a few drops of insanity"? -- <k>
-