Distributed detection over wireless sensor networks using sparse active sensors estimation

Abstract

Next-generation wireless networks require a large number of sensing devices for distributed communications with a simple design for wireless sensor networks (WSNs) and Internet of Things (IoT) applications [1]. In distributed communications, a large number of sensor devices are sensing the event and communicating their decision to a central node, i.e. fusion center. The impact of a number of sensor nodes, the number of antennas at the fusion center and noise distribution are analyzed on the global false alarm and miss detection rate for a WSN in the literature [1, 2]. However, all the sensor nodes are not active1 at a time in WSN system, which results in a sub-optimal system’s performance due to improper resource distribution such as system energy and channel bandwidth. Therefore, a system’s performance can be optimized by estimating the active node before the final detection or estimation at the fusion center. After estimating the active nodes in a system, resource distribution can be optimized and results in better performance. © 2019 URSI. All rights reserved.