Joint Sensor Scheduling and Target Tracking with Efficient Bayesian Optimisation

The received signal strength measurement has been widely used in search and tracking applications and its benefit is linked with the distance between the transmitter and receiver. This paper proposes an online Bayesian optimisation-based approach that relies on signal strength measurements to schedule multiple sensors for searching and tracking a moving target, without any prior knowledge of the target's state or motion model. A unique contribution lies in incorporating the Gaussian processes factorisation method into the Bayesian optimisation framework, which enhances the effectiveness of the proposed approach. Numerical results obtained from different sizes of measurements demonstrate that the proposed approach can efficiently schedule two unmanned aerial vehicles. Particularly, it achieves at most 21% lower computational time for deciding measurement locations and 79% lower time for updating the surrogate model as compared to the benchmark approach.