Target Tracking with Energy Efficiency Using Robotic Fish-based Sensor Networks

The achievements in underwater robotics and Wireless Sensor Networks make it possible to explore aquatic environment further. Target tracking which combines the mobility of underwater robotics with sensor networks has become an important application area in aquatic environment. This paper employs a Robotic Fish-based Sensor Network (RFSN) as mobile sensing platform to realize the target tracking. In order to guarantee the quality of the tracking, maintain the network connectivity, and consume as litter energy as possible, a novel approach for the target t racking is proposed. First, a locomotion model of fish is built by employing Lighthill's large-amplitude elongated-body theory and airfoil theory. Second, we define three functions with respect to the bias of the caudal fin based on the robotic fish dynamic characteristics and Kalman Filtering, which are the tracking quality function, the RFSN connectivity status function, and energy consumption function. Subsequently, the tracking problem can be transformed to a multi-objective optimization problem so that we get the individual control input of the fish which decides the fish how to swim. The simulation results show the effectiveness of our approach.