MASCOTTE no longer exists => visit the new project-team
Internship MASCOTTEMinimization of network power consumption with WAN Optimization
by Truong Khoa PHAN
| | | | Advisor | Joanna Moulierac | | School | Université de Nice Sophia Antipolis, Poytech'Nice | | Degree | Master IFI Parcours Ubinet | | Period | 01/03/11-31/08/11 | | Report | Available in pdf  |
Reducing energy consumption has become a key issue for industry, and therefore it is also an
important eld of networking research. Recently, many studies of energy-aware routing have been conducted.
The purpose of these works is to aggregate trac demands over a subset of the network links, allowing other
links to be turned o to save energy. Since energy-aware routing uses fewer number of links at any moment,
it is important to make sure that links are not overloaded. As a result, link capacity is the main constraint in
energy-aware routing problem. In this work, we propose a new energy-aware routing model with the support
of WAN Optimization Controller (WOC) - a device that can compress or eliminate redundant data trac.
Using theWOC, the capacity of network links is virtually increased, providing more space to aggregate trac
ows on the WOC-enabled links. We consider a simpli ed architecture that the router in the future can
integrate the WOC (called WOC-router). Based on this architecture, we propose algorithms for the three
following cases: (1) All routers on the network are the WOC-routers, we can enable or disable the WOC
in routers on demand; (2) There are some WOC-routers that have already been placed on the network, we
try to use both these WOC-routers and the normal routers on the network for trac aggregation; and (3)
There are a limited number of WOC-routers, hence we try to nd the best location for them on the network.
We rst formally de ne the problem and model it as an Integer Linear Program. Then, we propose greedy
heuristic algorithms and present simulation results on basic network (square grid) and on some real network
topologies. The results show that our algorithms achieve close to the optimal solution. In addition, our new
model can save more energy than the existing energy-aware routing without the WOC-routers.
List of interships | 
