Survey of IEEE 802.11 Wi-Fi deployments for deriving the spatial structure of opportunistic networks

An understanding of the underlying spatial structure of user-deployed wireless networks is invaluable for the design and optimization of opportunistic small-cell technologies in emerging multi-tier architectures. IEEE 802.11 Wi-Fi networks provide a real-world example of such a large-scale random network structure. In this paper we report on a high-resolution measurement survey of Wi-Fi deployments in Germany as a step towards deriving more sophisticated spatial models for transmitter distributions in emerging small-cell networks, which are the core of offloading strategies for future wireless Internet. Using a custom setup, our measurement campaign covered a range of urbanization and land use scenarios representative of Wi-Fi deployments in developed countries. Our data indicates a 14-fold increase in Wi-Fi density in urban residential areas over the last decade, consistent with increased broadband penetration. This supports our hypothesis that household density has become a strong predictor of access point (AP) density. We infer AP locations using a novel localization technique suitable for large measurement campaigns using off-the-shelf equipment. From this data, we derive pertinent spatial statistics to characterize the topologies of opportunistic networks. We derive nearest-neighbour distributions showing that realistic inter-neighbour distances are higher than predicted by a homogeneous Poisson Point Process, underlining the need for more refined spatial models than the random node location models which are currently prevalent. We also derive the two-pair correlation functions for the AP locations, showing that AP clustering is apparent over a large range of distances in our data set, which has implications for routing and interference mitigation.