Decision Fusion for Power-Constrained Wireless Body Sensor Networks with Amplify-and-Forward Relays

Abstract

This paper considers deriving the optimal decision fusion rule for cooperative wireless body sensor networks (WBSNs). The undertaken network model considers multiple sensors deployed to sense a certain binary biological phenomenon, and amplify-and-forward (AF) relays are deployed to assist the sensors to transmit their binary decisions to a remotely located fusion center (FC). The FC is responsible for processing the sensory decisions to provide a final global decision about the biological phenomenon. Since WBSN are subject to power constraints due to health issues and mobility enhancement, sensors and relays are assumed sharing a fixed power budget. The optimal fusion rule is derived using the likelihood ratio test while taking the AF relays and power constraint into account. Monte Carlo simulation is used to evaluate the detection and fusion error performance of the optimal rule, and compares them to the performance of two well-established suboptimal fusion rules for several operating conditions. The obtained results show that adding one AF relay for each sensor and using the optimal fusion rule can improve the probability of detection by about 5 dB for a wide range of signal-to-noise ratios (SNRs). Similarly, the fusion error probability may improve by about 10 dB. Moreover, the results show that the optimal rule significantly outperforms the suboptimal rules in several operating scenarios.