Analysis and optimisations in depth-based routing for underwater sensor networks

Underwater Sensor Networks (UWSNs) employ sensor nodes and acoustic communication to detect physical attributes of water such as temperature, pressure, etc. Research on UWSNs has emerged thanks to their wide spectrum of applications which includes the management of the oil reservoirs and the prevention of aqueous disasters, as well as military surveillance. The dynamic conditions of water, the energy constraints and the high error probability during data transmission are prominent challenges in the design of routing protocols in UWSNs. One of the main routing schemes is Depth-based routing (DBR) that performs a specialized anycast routing to the surface sinks, based along the depth measured from pressure sensors. In this thesis, we study and optimise some routing protocols for UWSNs, specifically those based on DBR. To this aim, we designed a novel simulator for studying DBR and its enhancements. Our simulator is based on AquaSim-NG and NS-3 (Network Simulator). With respect to the state of the art, we implemented the cross-layer communication required by DBR and an accurate representation of the operational modes of acoustic modems with the associated energy consumption. We developed some analytical models for UWSNs with the aim of a) identifying the optimal transmission range for sensor nodes given the state of the system, b) finding the optimal number of hops between the source and destination under various network settings, c) evaluating the role of the depth threshold in the definition of the routing scheme. In this work, a pivotal role is played by the energy consumption and expected lifetime of the network. Finally, based on our findings, we designed the Residual energy-Depth (RD) routing protocol which improves the network lifetime.