[88020] in North American Network Operators' Group
Intradomain Traffic Engineering
daemon@ATHENA.MIT.EDU (Hao Wang)
Wed Jan 18 00:06:52 2006
Date: Wed, 18 Jan 2006 00:06:24 -0500
From: Hao Wang <seraph.wang@gmail.com>
To: nanog@merit.edu
Errors-To: owner-nanog@merit.edu
------=_Part_59145_9311315.1137560784766
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline
Hi All,
I'm a PhD student currently studying intra-domain traffic engineering,
and I have two questions that I really wish to hear some opinions from you
network operators.
I'm experimenting with a prediction-based intra-domain traffic
engineering technique. The technique uses traffic demand matrices observed
in the history to predict future traffic demands, and computes a routing
that minimizes maximum link utilization (MLU) for those future demands.
I evaluate the performance of the technique using Abilene traffic trace=
s
collected at every 5 minutes interval. The results show that when the model
is able to predict the real traffic matrix, the technique can achieve close
to optimal MLU. However, when the model makes wrong prediction, the
technique suffers very high MLU (as high as 140%).
Basically, I have the following two questions:
1. In the traces I have, there exist several intervals with a huge,
sudden increase of traffic on some links. The prediction model I use cannot
predict those 'big spikes'. Do these 'big spikes' really happen in
operational networks? Or are they merely measurement errors? If they really
happen, is there a gradual ramp up of traffic in smaller time scale, say, o=
n
the order of tens of seconds? Or do these 'big spikes' really occur very
quickly, say, in a few seconds?
2. I have the option to make a tradeoff between average case performanc=
e
and worst case performance guarantee, but I don't know which one is deemed
more important by you. Are ISP networks currently optimized for worst case
or average case performance? Is the trade-off between these two an appealin=
g
idea, or may the ISP networks are already doing it?
I really appreciate any feedback from you about the above two questions=
,
and your help will be acknowledged in any publication about this work.
Thanks,
Edgar
------=_Part_59145_9311315.1137560784766
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline
<p>Hi All,<br> I'm a PhD student currently studying intra=
-domain traffic engineering, and I have two questions that I really wish to=
hear some opinions from you network operators.<br> I'm e=
xperimenting with a prediction-based intra-domain traffic engineering techn=
ique. The technique uses traffic demand matrices observed in the history to=
predict future traffic demands, and computes a routing that minimizes maxi=
mum link utilization (MLU) for those future demands.
<br> I evaluate the performance of the technique using Ab=
ilene traffic traces collected at every 5 minutes interval. The results sho=
w that when the model is able to predict the real traffic matrix, the techn=
ique can achieve close to optimal MLU. However, when the model makes wrong =
prediction, the technique suffers very high MLU (as high as 140%).
<br> Basically, I have the following two questions:<br>&n=
bsp; 1. In the traces I have, there exist several intervals wit=
h a huge, sudden increase of traffic on some links. The prediction model I =
use cannot predict those 'big spikes'. Do these 'big spikes' really happen =
in operational networks? Or are they merely measurement errors? If they rea=
lly happen, is there a gradual ramp up of traffic in smaller time scale, sa=
y, on the order of tens of seconds? Or do these 'big spikes' really occur v=
ery quickly, say, in a few seconds?
<br> 2. I have the option to make a tradeoff between aver=
age case performance and worst case performance guarantee, but I don't know=
which one is deemed more important by you. Are ISP networks currently opti=
mized for worst case or average case performance? Is the trade-off between =
these two an appealing idea, or may the ISP networks are already doing it?
<br> I really appreciate any feedback from you about the =
above two questions, and your help will be acknowledged in any publica=
tion about this work.</p>
<p>Thanks,<br>Edgar</p>
------=_Part_59145_9311315.1137560784766--
