An Efficient Localization Scheme With Velocity Prediction for Large-Scale Underwater Acoustic Sensor Networks

Abstract

Localization is vital and fundamental for underwater acoustic sensor networks (UASNs), as it provides location information for UASNs to achieve various practical underwater tasks. Most existing localization methods assume small-scale scenarios without battery energy constraints, making it inapplicable to large-scale UASNs. In large-scale UASNs, localization suffers from the challenges of excessive energy consumption and large localization error because of harsh underwater conditions like node mobility and huge ranging errors. To this end, we propose an efficient localization scheme with velocity prediction (LSVP) to solve the above challenges for large-scale UASNs. LSVP considers node mobility, ranging errors, and energy balance in a unified framework, which is applicable to realistic and scalable UASNs. Specifically, we first design a Doppler-assisted velocity prediction (DVP) algorithm to decrease energy consumption, which can solve the excessive communications caused by node mobility under ocean currents. Then, a acrlong CIL algorithm is proposed to decrease the localization error, which can reduce location uncertainty and error propagation caused by ranging errors. Extensive simulation results indicate that LSVP can achieve accurate velocity prediction and high precision localization for large-scale UASNs.