Collaborative Energy-Efficient Moving in Internet of Things: Genetic Fuzzy Tree Versus Neural Networks

The sensing application of space surveillance has put forward challenges to the Internet of Things (IoT). However, current moving algorithms in IoT rarely aim for target surveillance. In view of energy efficiency for multimodal signals in IoT, this paper mainly investigate three typical target trajectories: 1) line; 2) square; and 3) circle. On a basis of target learning, two types of collaborative sensor movement algorithms are proposed and compared. One approach is based on genetic fuzzy tree (GFT) and the other is based on the neural network (NN). Both algorithms can balance the energy consumption and the tracking performance. Simulation results show that the GFT-based algorithm outperforms NN-based algorithm in tracking error, but it demands more computational cost than that of NN-based scheme. This important result can provide intellectual sensing support in IoT applications, such as target surveillance, anti-terrorism, and unmanned border awareness.