Distributed Networked Sensing for Ocean Current Field Based on Grid Topology

Distributed networked sensing (DNS) has been applied to measure the ocean current velocity field. The features of the DNS of the ocean current field include: (i) a computationally efficient signal processing algorithm compared with the conventional tomographic inversion, (ii) a low system cost, and (iii) a potential for real-time monitoring of the ocean. For the signal processing method, a triangle with three sensor nodes is usually the basic processing unit in the literature. In this paper, we propose a square division (SD) based signal processing method for the DNS of the ocean current field. Error analysis will be given and compared for the proposed SD method and the widely used triangle-division (TD) method. A simulated ocean current field is constructed by the time-varying stream function based on a two-dimensional (2D) vortex model. The simulations show that the SD method has better performance than the TD method while having a lower network sparsity, defined as the number of acoustic paths per unit area. A sea trial is conducted to validate the core idea of the SD method, which shows that the estimated velocities agree with the direct measurements by the acoustic Doppler current profiler.