Dan Bridges wrote in a message to David Bowerman:
 DB> Thanks, it appears the reason I thought MNP4 was more efficient was
 DB> due solely to its typically larger frame size. When USR uses its
 DB> 244+7 byte frame there would not be any apparent difference. 
 DB> How did you derive these percentages? For example with (128+7) I
 DB> would assume the frame-support overhead would be 7/(128+7) = 5.2% 
The MNP and LAPM stuctures also add in a 0 bit on the average every 60 bits 
for transparency -- check a reference on SDLC/HDLC synchronous data 
transmission for more information on this.  The formula becomes 1 - ((packet 
size/(packet size+7)) * (60/61)).  So you need to use 1 - ((128/(128+7) * 
(60/61)) or 1 - 0.9326 = 6.74%.
Regards,
       David
--- timEd/2 1.10+
---------------