Underwater localization with sound velocity profile

Underwater acoustic sensor networks (UASNs) is a widely used enabling technique for various underwater applications. To facilitate these real applications, it is essential to obtain the position information of sensor nodes by localization methods. One of the most common localization approaches employs time-of-arrival (TOA) to measure and utilize the distances from the target node to several anchor nodes with known positions by simply multiplying the sound speed and propagation time. These existing works ideally assume a linear trajectory of acoustic wave in underwater. In practice, however, underwater sound velocity varies with temperature, salinity and pressure as depicted in sound velocity profile (SVP), which deteriorates the localization accuracy. In this work, we consider the underwater localization in the presence of SVP. By assuming that the SVP is roughly linear and the sound velocity only depends on the depth, we propose to apply particle swarm optimization (PSO) to further improve the localization accuracy. Numerical results suggest that the proposed algorithm achieves better localization performance than traditional approaches.