[e2e] Bandwidth Estimation

Cottrell, Les cottrell at SLAC.Stanford.EDU
Fri Oct 3 08:36:11 PDT 2003


I would be quite happy to change over to using bitrate for what I am currently using bandwidth for when writing about estimating "bandwidth" in the context of packet pair dispersion etc. Maybe this could be a start on getting some harmony going.

-----Original Message-----
From: J. Noel Chiappa [mailto:jnc at ginger.lcs.mit.edu] 
Sent: Friday, October 03, 2003 7:42 AM
To: end2end-interest at postel.org
Subject: Re: [e2e] Bandwidth Estimation


    > From: "David P. Reed" <dpreed at reed.com>

    > It's very sad to see that the vernacular use of "bandwidth" to mean
    > "information rate" has been adopted by the networking community.

Let me reach back into the stream and try and make a useful, practical suggestion, as opposed to pontificating about terminology drift in general..


    > there is no word or even an easy phrase for what "bandwidth" was coined
    > to mean - which is the range of frequencies occupied by a modulated
    > signal. ... impoverished by this theft of a term, because for us, it is
    > important to distinguish quite clearly between the information rate of
    > a signal and its bandwidth in our work.
    > ...
    > we probably have to live with this loss of a word.

Yes, I think "bandwidth" is a lost cause, alas. I suggest we come up with two new words, "bitrate" (which is nice and short and snappy, and accurate to boot), and something new for what "bandwidth" used to be, and focus on using them (and hooting whenever anyone says "bandwidth" :-).

    > So I suppose we'll have to coin a new word for the important old
    > meaning. I have proposed "frequespan".

I've been musing about this for a couple of days. My first thought was to go back to "band-pass", which still has its original meaning, but a) I don't want to lose it for *its* original use (in filters), and b) I don't think it's as self-explanatory as it should be.

I've tried a number of possibilities out in my mind, and the one I like best is "band-size", which I think is better than "frequespan" (and not just because I suggested it :-). It's euphonious, to-the-point, and says just what it means - the size of the frequency range (band). What do you think?


    > Bandwidth is measured in Hertz. Information rate (or bitrate) is
    > typically measured in bits/second (and I am considering a formal
    > proposal to IEEE to call for a new unit, the Shannon, to deal with the
    > lack of a named unit).

Hear, hear! Let us know if we need to get up any kind of support to help you with this. This also fits very well with "bit-rate".

    > Of course, the voracious ignorance of CS professionals will probably
    > colonize that one as well ... 

Based on past experience, I'm actually hopeful. I think this discussion has helped to bring this to the surface in the community, and if the outcome is something with an even wider distribution (e.g. a short RFC, or an announcement to the IETF list), that should really start the ball rolling well enough, I think.

	Noel




More information about the end2end-interest mailing list