Distributed homology algorithm to detect topological events via wireless sensor networks

Wireless sensor networks (WSNs) can span large geographical regions and collaboratively monitor environmental phenomena, for example, forest fires. By designing a WSN to detect changes to such phenomena, current environmental monitoring systems could be supplemented, if not replaced. This research focuses on incremental insertion events, arising from the elevation of a single node´s sensor reading between two consecutive states of network operation. Homology, a field of topology, is used to detect and differentiate between incremental insertion events of interest. Particular homology tools are translated to a distributed environment for 2-dimensional WSN deployments. The result is a novel distributed algorithm that can compute an incremental insertion event associated with a region comprising n nodes in O(n) time, using O(n) storage, and O(n) data passed via messages. A small-scale, laboratory testbed is developed to evaluate the algorithm. Deployment results indicate that only nodes in physical proximity to an event are tasked, thereby conserving network resources and allowing multiple disparate events to be simultaneously monitored. Further, transmission cost is shown to vary linearly with the size of the evolving region, confirming one component of the formal analysis.