Re: Error in cwin estimation?

From: Mark Allman (mallman@grc.nasa.gov)
Date: 04/09/01

  • Next message: Mariano Korman: "What about UDP?"

    Message-Id: <200104091924.PAA09613@guns.lerc.nasa.gov>
    From: Mark Allman <mallman@grc.nasa.gov>
    Subject: Re: Error in cwin estimation? 
    Date: Mon, 09 Apr 2001 15:24:32 -0400
    
    

    > I am glad you raised the issue, because, if the monitor is 'near'
    > the receiver, there is no way to determine cwin, and ownd will be,
    > in most of the cases, 2*MSS, due to delayed ACKs from the
    > receiver. Is there any way to go around this and determine the
    > actual cwin?

    Maybe... There is an algorithm in the following paper that tracks
    the segments that are liberated by a given ACK during slow start.

        Mark Allman, Vern Paxson. On Estimating End-to-End Network Path
        Properties. ACM SIGCOMM, September 1999, Cambridge, MA.
        http://roland.grc.nasa.gov/~mallman/papers/estimation.ps

    My *guess* is that with some elbow grease you might be able to use
    that during both slow start and congestion avoidance to estimate the
    amount of outstanding data. But, that is just a guess.... I don't
    think it is a trivial extension.

    > I presume you meant
    > ownd += ownd_sample * (time_since_last_sample / transfer_time)

    Yep.

    > It gives a nice weighting, therefore more insight, but I don't
    > know if it is necessary...After all, when, for example, average
    > RTT is determined, it is not weighted with the above term...

    True.

    However, the RTT, unlike the cwnd, is a property of the network
    path. So, we can't assume that the RTT is static between samples as
    the cwnd is.

    Maybe just two ways of looking at things...

    allman

    ---
    Mark Allman -- BBN/NASA GRC -- http://roland.grc.nasa.gov/~mallman/
    ----------------------------------------------------------------------------
    To unsubscribe, send a message with body containing "unsubscribe tcptrace" to
    majordomo@tcptrace.org.
    



    This archive was generated by hypermail 2b30 : 04/10/01 EDT