[e2e] useful statistics using RTT

[uaca at alumni.uv.es]

Hi all

I'm doing measurements of RTT on a WAN interface

Currently I'm not storing the measurements itself but the

- sum(RTT)   --> the sum of all measurements
- sum(RTT^2) --> the sum of the square of the measurements
- N          --> the number of measurements made

so I can calculate, mean and standard variation for a given period. 

I currently make these statistics for two sets of networks

- the RTT between the network where the WAN interface belongs and the rest
  of the autonomous system wich the network belongs
- the RTT between the network where the WAN interface belong and the rest
  of the Internet.

I have to make this big two sets becuase otherwise I will no have enough samples
for each set to make statistics significant.

I would appreciate comments and suggestions about what statistics/methology
would be useful to apply to this data. 

In short, what you would do if you have such data?

maybe an ewma (or any other quality control) procedure would be
useful/reasonable?

Thanks in advance

	Ulisses

                Debian GNU/Linux: a dream come true
-----------------------------------------------------------------------------
"Computers are useless. They can only give answers."            Pablo Picasso

--->	Visita http://www.valux.org/ para saber acerca de la	<---
--->	Asociación Valenciana de Usuarios de Linux		<---