Re: Another Point Of View (Round 2)

[Bill Frezza <frezza@interramp.com>]

  At 10:33 PM +0000 11/29/96, Mike Bilow wrote:
  >As I
  >think should be obvious to anyone who made it past high school math, the only
  >real issue is whether the value of d is different for voice calls and data
  >calls.
  
  No, this is not correct. This false assumption is the root of your error.
  The two probability density functions have completely different forms.
  Poisson distributions have much smaller standard deviations than Power Law
  distributions even if the means are equal. (You can see this just by
  looking at histograms of raw call data.) The "flatter" nature of measured
  internet call holding-time distribution functions includes many more long
  holding-time calls than would be produced by an exponential distribution.
  Most significantly, these longer holding-time calls have a disproportionate
  impact on total call-minutes, which is one of the two real load issues you
  correctly identified. The second is the overall Quality of Service impact
  of having large hunt groups tied up with 30+ CCS of traffic.
  
  I'll try to run down a current technical reference on the PDFs if you're
  really interested in understanding this, it's been a long time since my
  traffic modelling days back at Bell Labs.
  
  Bill Frezza