[e2e] Random losses on very high speed networks

Cannara cannara at attglobal.net
Tue Jul 22 22:07:50 PDT 2003


Constantine, I was not bashing, just observing and opining, as you are. 
You're free to misuse any part of the language.  Anyone is.  They just needn't
command respect for doing so needlessly.

As an electrical engineer myself, even a PhD from Stanford, I may even dain to
comment on your remarks 1) & 2) below -- signal & channel bandwidth are
measured in the same way, in the same units, with the same meaning, just as
any signal and filter algorithm are mutually transformable.  I'm not arguing
with you, just reviewing facts so that those without specific training in this
area who are listening in aren't misled.

On "overloading", a rather cute term itself from lazy interface design, it's
rarely a good idea, and in the case of bandwidth/capacity, totally unnecessary
scientifically.

Alex

Constantine Dovrolis wrote:
> 
> Alex, "bandwidth" is certainly not the first technical term
> that is being overloaded, even in the scientific/research
> literature, with multiple meanings.
> 
> Even within the context of its "physical" interpretation (Hz),
> the term "bandwidth" is used to describe two different things:
> 
> 1. We talk about the "bandwidth of a signal", defined based
> on the range of a signal's spectrum that carries most of the
> signal's energy.
> 
> 2. We also talk about the "bandwidth of a channel", or system
> in general, defined by the difference of the cutoff frequencies
> in the system's transfer function.
> 
> There is nothing wrong with overloading a term, as long as
> it is clear from the context which concept you refer to.
> 
> Of course I will not reply to the bashing part of your message..
> 
> Constantinos
> 
> --------------------------------------------------------------
> Constantinos Dovrolis | 218 GCATT | 404-385-4205
> Assistant Professor | Networking and Telecommunications Group
> College of Computing | Georgia Institute of Technology
> dovrolis at cc.gatech.edu
> http://www.cc.gatech.edu/fac/Constantinos.Dovrolis/
> 
> On Tue, 22 Jul 2003, Cannara wrote:
> 
> > I don't believe anyone was whining Constantine.  It's simply that bandwidth is
> > the wrong word to use in English for data rate, capacity, etc.  It remains
> > wrong no matter who uses, or used it, and for how long.  In an engineering
> > sense it's a wrongful confusing element in communication with others, so in
> > one sense its misuse is bad sientific manners, especially for text read by
> > people new to the discipline.
> >
> > In fact, it clouds the scientific background of research that has gone on,
> > since before Shannon and Nyquist, into symbol transmission.  It's hard to
> > justify misuse of the term when explaining how much effort has gone into
> > optimal signal encoding over many decades, in particular the amazing efforts
> > by modem designers, including Hayes.  No technical discussion of how modem
> > designers plowed the fertile field of bandwidth to increase capacity over the
> > years would make sense if bandwidth equalled capacity.
> >
> > Personally, I find it humorous to see how many folks who advertise themselves
> > as learned in the field use "bandwidth" so incorrectly and persistently --
> > almost as if it's PC, and they're afraid to rock anyone's boat from whom they
> > may need funding or some other support in the future. It's an intersting
> > social phenomenon to me, becuase it demonstrates how some folks will sacrifice
> > even scientific accuracy for personal comfort. In any case, my students don't
> > misuse the terms on homework & tests.  :]
> >
> > Alex
> >
> > Constantine Dovrolis wrote:
> > >
> > > > > I agree with you that "bandwidth" is largely misused in the
> > > > > networking and CS community -- anyone who takes a networking class
> > > > > and still misuses bandwidth for capacity (at any layer) shoulda
> > > > > flunked.  :]
> > > >
> > >
> > > Folks, frankly, it is way too late to whine about
> > > the use of the term "bandwidth" in the networking context. People
> > > have been using this term to refer to "capacity or throughput" (bps),
> > > as well as to "physical bandwidth" (Hz), for decades. I can
> > > recall the former use of the term in Jacobson's 88 paper, but I am
> > > sure that there are earlier references too.
> > >
> > > Besides, the context in which the term is used is so different
> > > for Hz vs bps that probably it never causes confusion. And at the
> > > end of the day, the two "bandwidth" concepts are not totally
> > > irrelevant after all - see Shannon's law.
> > >
> > > Constantinos
> > >
> > > --------------------------------------------------------------
> > > Constantinos Dovrolis | 218 GCATT | 404-385-4205
> > > Assistant Professor | Networking and Telecommunications Group
> > > College of Computing | Georgia Institute of Technology
> > > dovrolis at cc.gatech.edu
> > > http://www.cc.gatech.edu/fac/Constantinos.Dovrolis/
> >
> >




More information about the end2end-interest mailing list