Topology control and manipulation in a large scale wireless network testbed
Stefan Bouckaert, Bart Jooris · Ghent University Academic Bibliography (Ghent University) · 2009
While analytical models and well-considered network simulations are invaluable for wireless network researchers, they cannot capture the full complexity and dynamics of the error prone wireless medium. The simplified assumptions that are necessarily used in simulations and calculations often cause unwanted side effects, causing protocols to fail or perform considerably worse than expected under real-life conditions. Consequently, the international research community is increasingly recognizing the added value of experimental wireless network deployments. Recently, the w-iLab.t testbed was installed at the three floors of the buildings of the IBCN research group and at the offices of the IBBT research institute in Ghent, Belgium. With nearly 200 wireless WiFi nodes, it is one of the largest wireless testbeds of its kind installed in a real office environment worldwide. As the output power of the wireless interfaces cannot be set to an arbitrarily low value, and the area of the test site is limited, all nodes are within each other’s transmission range by default. Therefore, several techniques to manipulate the topology in testbeds are briefly discussed in this paper. It is also shown how by combining techniques, the topology of the wiLab.t can be manipulated in order to recreate both very sparse and very dense networks.