Hierarchical censoring for distributed detection in wireless sensor networks

In energy-limited wireless sensor networks, detection using 'censoring sensors' reduces the probability that a sensor must transmit, thereby saving energy. In this paper, we introduce a hierarchical distributed detection scheme designed specifically for multihop networks. If a sensor's local likelihood ratio (LLR) crosses a threshold, it is sent to the next higher level sensor. A simple feedback scheme is also considered. We study the performance of a Gaussian change-of-mean detection system using this hierarchical censoring scheme, with and without feedback. We show that good detection performance can be achieved while significantly reducing sensor transmissions compared to the optimal detection system.