Multiple Hypothesis Testing Framework for Spatial Signals

The problem of identifying regions spatially interesting, different or adversarial behavior is inherent to many practical applications involving distributed multisensor systems. In this work, we develop a general framework stemming from multiple hypothesis testing identify such regions. A discrete spatial grid assumed for the monitored environment. points associated with hypotheses are identified while controlling false discovery rate at pre-specified level. Measurements acquired using large-scale sensor network. We propose novel, data-driven method estimate local rates based on spectral moments. Our agnostic specific propagation models underlying physical phenomenon. It relies broadly applicable density model summary statistics. between sensors, locations assigned interpolated rates. benefits our illustrated by propagating radio waves.