Understanding wireless network coverage, connectivity, and locations using opportunistic sensing for a WiFi advisory system

As crowdsourcing using sensor-enabled smartphones boosted the confluence of human mobility and ambient network information, it opens up more opportunities for many promising crowd-centric application systems. This paper considers a WiFi advisory system, where crowdsourcing technology is used to gather information of wireless networks for understanding wireless network capability better including the coverage, connectivity, and locations of wireless networks. Such a system will address two critical challenges: (1) no self-contained bootstrapping data and (2) device-dependent data processing. To address the first one issue, we design and implement an opportunistic sensing system for smartphones to collect wireless information along human mobility without human engagement. To address the second issue, we design a data aggregation algorithm to infer the locations of wireless networks based on crowdsourced data from several different smartphones. In addition, we also analyze the computation complexity of our data aggregation algorithm. Finally, we implement our wireless network monitoring infrastructure to study the spatial and temporal variation and patterns of wireless networks systematically.