Loading ...
Sorry, an error occurred while loading the content.

Re: [BPQ32] Winmor speed?

Expand Messages
  • Robert S. Fisher
    It s the overhead at the start of the connection. Winmor does a lot mode shifting with poor paths at the first of the connect especially if the path is poor
    Message 1 of 3 , Mar 4, 2010
    • 0 Attachment
      It's the overhead at the start of the connection. Winmor does a lot mode
      shifting with poor paths at the first of the connect especially if the
      path is poor and the connect starts at 1600BW. It will mode shift until
      it finds a suitable modulation mode. With larger transfers it has more
      time for the Bytes/min average to increase. In testing that I have done
      on paths that are expected to be poorer and with smaller transfers, it
      actually can be a quicker session if it starts in 500BW.

      Rob


      kt4wo67@... wrote:
      >
      > Let me say this 1st:
      > I have ONLY used BPQ/Winmor to pull nodes/routes/ports,etc lists from
      > a few other users and send one message.
      >
      > Has there been any independent speed tests on winmor?
      >
      > After linking and pulling some lists last night, I went
      > KtoK on ALE400... and it "seemed" faster,,,alot faster.
      > I know the band, station, etc play a large part.
      > I have been on digi modes for over 15 years...so I know this.
      > I have had some good links with KB8OAK-7 and I dont know
      > if its just the overhead with small transfers or my imagination (and
      > it very well may be) but I am not seeing the speeds that I have read
      > about.
      >
      > what do ya'll think?
      >
      > Trip - KT4WO
      >
      >
    • Rick Muething
      Trip, The speed (net bytes/min) in WINMOR is a function of: The modulation mode used 4FSK, 4PSK, 8PSK, 16PSK, And the number of repeats necessary ( for
      Message 2 of 3 , Mar 5, 2010
      • 0 Attachment

        Trip,

        The “speed” (net bytes/min) in WINMOR is a function of:

        The modulation mode used 4FSK,  4PSK,  8PSK,  16PSK, 

        And the number of repeats necessary ( for ARQ operation)

         

        For FEC (non ARQ) Keyboard/broadcast mode speed is fixed speed using Viterbi 4FSK, the slowest but most robust mode.  The speed there is either 391 bytes/minute for the 500 Hz wide mode or 1565 byte/min for the 1600 Hz mode.  Since non ARQ modes can’t insure error free delivery there is no compression used on text in non ARQ.

         

        For ARQ operation, which WINMOR was really designed for, the speed before considering repeats but including ARQ overhead ranges:

        503 bytes/min (500 Hz mode) or 2012 bytes/min (1600 Hz mode) using the slowest and most robust mode 4FSK

        To

        2465 bytes/min (500 Hz mode) or 9856 bytes/min (1600 Hz mode) for the fastest and least robust mode 16PSK

         

        Since ARQ does deliver error free (after ARQ and FEC correction) bytes compression of text is possible and used in messages and that results in about a 80-100% average improvement in speed on a typical 2K text message. [You would approximately double the bytes/min when using ARQ to get Characters per minute]

          

        During ARQ operation the “gear shift algorithm” constantly monitors the throughput and adjusts the modulation mode to try and maximize the throughput for the current channel propagation. In practice a good channel with 10-15 db S/N and minimal multipath will generally yield about 70-80% of the max ARQ throughput above.

         

        The numbers above all pulled directly off the WINMOR rate sheet posted on the WINMOR reflector  http://groups.yahoo.com/group/WINMOR

         

        During operation RMS Express displays the rolling 1 minute average throughput and summarizes the session throughput stats at the end of a session.

         

        To be able to make any real apples to apples comparison between modes it is necessary to run the tests in a controlled and repeatable HF channel such as provided by a channel simulator. This allows consistency and accuracy in the measurements.

         

        It is also necessary to distinguish between ARQ and non ARQ modes.  They are different animals and usually used for different types of communication. 

         

        I hope in the next few months to be able to run controlled experiments on several modes using the HF simulator and automatic data collection. These tests take a long time because you generally must have runs of 10 minutes or more at each channel condition and mode to collect accurate statistics even with the HF simulator. 

         

        Rick KN6KB

         


        From: BPQ32@yahoogroups.com [mailto:BPQ32@yahoogroups.com] On Behalf Of kt4wo67@...
        Sent: Thursday, March 04, 2010 10:12 PM
        To: BPQ32@yahoogroups.com
        Subject: [BPQ32] Winmor speed?

         

         

        Let me say this 1st:
        I have ONLY used BPQ/Winmor to pull nodes/routes/ ports,etc lists from
        a few other users and send one message.

        Has there been any independent speed tests on winmor?

        After linking and pulling some lists last night, I went
        KtoK on ALE400... and it "seemed" faster,,,alot faster.
        I know the band, station, etc play a large part.
        I have been on digi modes for over 15 years...so I know this.
        I have had some good links with KB8OAK-7 and I dont know
        if its just the overhead with small transfers or my imagination (and it very well may be) but I am not seeing the speeds that I have read about.

        what do ya'll think?

        Trip - KT4WO

        No virus found in this incoming message.
        Checked by AVG - www.avg.com
        Version: 9.0.733 / Virus Database: 271.1.1/2707 - Release Date: 03/03/10 14:34:00

      Your message has been successfully submitted and would be delivered to recipients shortly.