Algorithm 2 Detection of regimes in round-trip time sequences
function DetectRegimes(X,$\delta$,s)
$X=\langle x_1,x_2,\dots ,x_n\rangle$ is the scenario, $\delta$ the distribution, $s>0$ the min. regime length
$R\leftarrow \varnothing$	▹ list of regimes detected
$l\leftarrow \varnothing$	▹ last model (regime) found
$f_0\leftarrow 1$	▹ start index of the last model within X
for $f\in \langle 1,n\rangle$ do
if $f-f_0+1\ge s$ then
$m\leftarrow AssessModel(\langle x_{f_0},x_f\rangle ,\delta )$	▹ MLE + GoF of segment $[f_0,f]$ with distrib. $\delta$
if $m=\varnothing$ then	▹ failed model with the new RTT $x_f$
if $l\ne \varnothing$ then	▹ last model ended here
$R\leftarrow R\cup (f_0,f,l)$	▹ new regime is built with last valid model
$l\leftarrow \varnothing$	▹ no last model
$f_0\leftarrow f$	▹$x_f$ has not been used for building last regime
else	▹ no previous model; slide the window forward
$f_0\leftarrow f_0+1$
end if
else	▹ model assessed ok; go on extending it
$l\leftarrow m$
end if
end if
end for
if $l\ne \varnothing$ then	▹ trailing model
$R\leftarrow R\cup (f_0,n,l)$
end if
return R
end function