A lot of attention has been given to multihop wireless networks lately, but further research—in particular, through experimentation—is needed. This attention has motivated an increase in the number of 802.11-based deployments, both indoor and outdoor. These testbeds, which require a significant amount of resources during both deployment and maintenance, are used to run measurements in order to analyze and understand the limitation and differences between analytical or simulation-based figures and the results from real-life experimentation. This paper makes two major contributions: (i) first, we describe a novel wireless multihop testbed, which we name FloorNet, that is deployed and operated under the false floor of a lab in our Computer Science building. This false floor provides a strong physical protection that prevents disconnections or misplacements, as well as radio shielding (to some extent) thanks to the false floor panels—this later feature is assessed through experimentation; (ii) second, by running exhaustive and controlled experiments we are able to analyze the performance limits of commercial off-the-shelf hardware, as well as to derive practical design criteria for the deployment and configuration of mesh networks. These results both provide valuable insights of wireless multihop performance and prove that FloorNet constitutes a valuable asset to research on wireless mesh networks.
Available at: http://works.bepress.com/michael_zink/1/