A Multimodal Data Harness Approach of Mobile Sensors Trajectory Planning for Target Tracking

The future Internet comprising multimodal mobile wireless sensors (MMWSs) dramatically increases the data dimension and complexity compared to the traditional homogeneous networks. It consists of various types of nodes that can be deployed in different spatial domains to form the space–air–ground–ocean-integrated network (SAGOI-Net). A key challenge to Internet of Things (IoT) with SAGOI-Net is harnessing the real-time data from these sensors to obtain target characteristics and realize target tracking, especially in rough environment with obstacles and enemy threats. In this article, we propose an approach—fuzzy logic and flocking control under an extended Kalman filter (EKFFL) to provide sensor nodes trajectory formation. With the tools from control systems and signal processing, this approach improves the commonly adopted flocking control algorithm and fully employs multi modality of data. Compared with the modified Kalman consistency filter (KCF) approach, the proposed EKFFL achieves higher target tracking accuracy with a shorter time period of trajectory formation. The tradeoff between the data delay and tracking accuracy is also analyzed. Based on its advantages, the EKFFL approach can be applied in broad scenarios, such as unmanned border awareness systems, environmental monitoring, and intelligent transportation.