Distributed detection with censoring of sensors in Rayleigh faded channel

This paper considers the problem of fusing decisions in a distributed detection system when the local binary decisions made at the sensors, relative to observations of a common binary phenomenon, are transmitted over Rayleigh faded channel subject to additive noise. We use a training-based channel estimator at the fusion center (FC) to estimate the complex Gaussian fading coefficients characterizing the channels between the sensors and the FC. Channel state information (CSI) is used on the fading coefficients for censoring the sensors. Locally optimal decision threshold is considered for binary quantization at the sensors. The detection error probability is selected as a qualitative measure of system performance in the presence of majority logic fusion at the FC and is evaluated by means of simulations. We study the effects of the channel estimation error, the channel signal-to-noise ratio (SNR), the sensor SNR, and the number of selected sensors on the system performance.