Weighted decision fusion vs. counting rule over wireless sensor networks: A realistic comparison

Distributed detection (DD) is one of the most important applications of wireless sensor networks (WSN). In DD, each sensor decides about an event occurrence and sends its decision to a fusion center (FC) where the final decision is made using a fusion rule. An intuitive fusion rule is counting rule fusion (CRF) in which the number of sensors which detect the event is compared with a threshold. Counting rule is proposed to counteract the practical issues of centralized detection scheme which is theoretically optimal but impractical in a large-scale network. In this paper, a weighted decision fusion (WDF) scheme is proposed in which decisions of sensors are weighted based on their estimated SNR values. In addition, each sensor which detects the event sends data to the FC if its received SNR is more than a pre-specified value. Detection performances of WDF and CRF are compared using a practical dataset.