Abstract

Suppose we have a packet network which we wish to make measurements on through active probing; we wish to determine the optimal probing rate which allows us to find the most information about the system, whilst simultaneously minimising the interference caused to the network by probing.

We use the statistical principles of design of experiments to regard probing as a numerical experiment that can be designed optimally. We demonstrate how to analyse the evolution of the network by modelling the system as a Markov Chain, and deduce its likelihood function, and hence the Fisher information matrix. From this, numerical results provide a guide to the best design for the experiment (probing rate) for different values of input parameters. We further develop our ideas to show what happens when we take into account the effect of the probes interfering with the experiment. We present examples, and demonstrate how this could be useful to determine optimal probing rates practically.