Underwater Target Tracking While Decreasing Energy Consumption in Bearing-Only Sonar Networks

Target tracking in obstacle-cluttered underwater environments is an important scenario of underwater sensor networks. This study assumes that bearing-only sonar sensors are deployed in underwater environments, so that they can measure the sound of a nearby target. Every bearing-only sonar sensor measures the bearing of a target, which exists within the sensing range of the sensor. This article assumes that we can access the obstacle information, as well as global location of each bearing-only sonar sensor. The goal of this study is to track a target, by integrating the bearing measurements of each sonar sensor. We consider the case where there exists a remote computation center where a target tracking is performed by integrating the bearing measurements of sonar sensors. We assume that wired communication links between a sensor and the computation center are available. Measuring the target sound consumes the power of each sonar sensor. Only in the case where a sonar sensor is awaken, the sensor measures a target’s sound direction and sends the measurement information to the computation center. In practice, the sensor network performance may be disturbed by communication bottleneck situations, as too many sensor nodes send the measurement information to the computation center at the same time. For tracking a target while decreasing the energy consumption and communication bottleneck, we propose a multi-sensor target tracking scheme that awakens only a sensor which has high probability of target detection. To the best of our knowledge, this study is novel in addressing an energy-saving target tracking based on bearing-only sonar networks in obstacle-cluttered underwater environments. Computer simulations prove that employing our multi-sensor tracking approach improves the estimate accuracy and increases the lifetime of the networks, compared to the case where the proposed approach isn’t employed.