Potential Function-Based Control Using Neutral Networks in Uncertain WSNs With Mobile Collectors

In wireless sensor networks (WSNs), sink node mobility has been considered an efficient way to alleviate excessive data forwarding at nodes close to the sink node. Since the mobile sinks (or collectors) move around the network collecting data from the sensor nodes, the need for multi-hop forwarding of sensor data is reduced. In WSNs with mobile collectors, efficient path planning of mobile collectors is a major challenge where many algorithms have been proposed. However, they do not consider tracking performance degradation during flight due to environmental factors such as wind, waves, and obstacles. Tracking control algorithms have been proposed to correct the paths of mobile collectors, but existing work do not consider the possibility of collisions between multiple collectors. In this paper, we propose a distributed control approach that can improve tracking performance of mobile collectors while avoiding collisions in unknown and uncertain WSN environments. The proposed algorithm uses a potential field of attraction and repulsion to control individual mobile collector input, which enables mobile collectors to adjust their trajectories with online learning in a distributed fashion. From the theoretical analysis, the upper bound of tracking error and the conditions of the system parameters to guarantee the desired tracking error set by the operator are derived. The simulation results show that the proposed method significantly outperforms existing methods in terms of tracking error, collision rate, system stability, and convergence.