Efficient Node Collaboration for Mobile Multi-Target Tracking Using Two-Tier Wireless Camera Sensor Networks

To address the Mobile Multiple Targets Tracking (MMTT) problem, we propose the two-tier Wireless Camera Sensor Network (WCSN) and the corresponding the sensor collaboration scheme, which is developed by the Cardinality Balanced Multi-Bernoulli (CBMeMBer) filtering algorithm. In our proposed scheme, at each time step the tracked targets are regrouped, and the sensor with the best view of each group of the tracked targets is selected as the new Cluster Head (CH), which activates the sensors located within their communication ranges to be the cluster members. Furthermore, we also extend the single-directional WCSN to multiple-directional WCSN to improve the target-location estimate accuracy. Our simulation results validate the performances of our proposed dynamic node collaboration scheme.