Learning to Communicate Underwater: An Exploration of Limited Mobility Agents

Abstract

Underwater environmental parameters vary with time and can negatively impact the quality of communication. The adaptive control system suggested attempts to improve communication in underwater networks where environmental conditions are stochastic and time-variant. Adaptive depth control is explored in limited mobility underwater acoustic sensor networks. The adaptive control seeks to leverage the acoustic properties along the thermocline sensors anchored to the seabed in order to improve link stability in underwater networks. The sensor is allowed two directions of movement, either surface or dive, in order to avoid physical phenomena that cause faults. The algorithm presented is capable of adapting to the optimal performance depth in unimodal stochastic stationary and non-stationary environments.