• Two-dimensional UW-ASNs for ocean bottom monitoring. These are constituted by sensor nodes that are anchored to the bottom of the ocean. Typical applications may be environmental monitoring, or monitoring of underwater plates in tectonics. 
• Three-dimensional UW-ASNs for ocean column monitoring. These include networks of sensors whose depth can be controlled, and may be used forsurveillance applications or monitoring of ocean phenomena (ocean bio-geo-chemical processes, water streams, pollution, etc). 
• Three-dimensional networks of Autonomous Underwater Vehicles (AUVs). These networks include fixed portions composed of anchored sensors and mobile portions constituted by autonomous vehicles.

Fig. 1 - Two-dimensional Underwater Sensor Networks

A reference architecture for two-dimensional underwater networks is shown in the figure above. A group of sensor nodes are anchored to the bottom of the ocean with deep ocean anchors. By means of wireless acoustic links, underwater sensor nodes  are interconnected to one or more underwater sinks (uw-sinks), which are network devices in charge of relaying data from the ocean bottom network to a surface station. To achieve this objective, uw-sinks are equipped with two acoustic transceivers, namely a vertical and a  horizontal transceiver. The horizontal transceiver is used by the uw-sink to communicate with the sensor nodes in order to: i) send commands and configuration data to the sensors (uw-sink to sensors); ii) collect monitored data (sensors to uw-sink). The vertical link is used by the uw-sinks to relay data to a surface station. Vertical transceivers must be long range transceivers for deep water applications as the ocean can be as deep as 10 km. The surface station is equipped with an acoustic transceiver that is able to handle multiple parallel communications with the deployed uw-sinks. It is also endowed with a long range RF and/or satellite transmitter to communicate with the onshore sink (os-sink) and/or to a  surface sink (s-sink).

Sensors can be connected to uw-sinks via direct links or through multi-hop paths. In the former case, each sensor directly sends the gathered data to the selected uw-sink. This is the simplest way to network sensors, but it may not be the most energy efficient, since the sink may be far from the node and the power necessary to transmit may decay with powers greater than two of the distance. Furthermore, direct links are very likely to reduce the network throughput because of increased acoustic interference due to high transmission power. In case of multi-hop paths, as in terrestrial sensor networks, the data produced by a source sensor is relayed by intermediate sensors until it reaches the uw-sink. This results in energy savings and increased network capacity, but increases the complexity of the routing functionality as well. In fact, every network device usually takes part in a collaborative process whose objective is to diffuse topology information such that efficient and loop free routing decisions can be made at each intermediate node. This process involves signaling and computation. Since, as discussed above, energy and capacity are precious resources in underwater environments, in UW-ASNs the objective is to deliver event features by exploiting multi-hop paths and minimizing the signaling overhead necessary to construct underwater paths at the same time.

A protocol stack for uw-sensors should combine power awareness and  management, and promote cooperation among the sensor nodes. It should consist of physical layer,  data link layer, network layer, transport layer, and  application layer functionalities.
The protocol stack should also include a power management plane, a coordination plane, and a localization plane. The power management plane is responsible for network functionalities aimed at minimizing the energy consumption (e.g., sleep modes, power control, etc.). The coordination plane is responsible for all functionalities that require coordination among sensors, (e.g., coordination of the sleep modes, data aggregation, 3D topology optimization, etc.). The localization plane is responsible for providing absolute or relative localization information to the sensor node, when needed by the protocol stack or by the application.

It is worth pointing out that while all the research on underwater networking so far has followed the traditional layered approach for network design, it is an increasingly accepted opinion in the wireless networking community that the improved network efficiency, especially in critical environments, can be obtained with cross-layer design approaches. These techniques will entail a joint design of different network functionalities, from modem design to MAC and routing, from channel coding and modulation to source compression and transport layer, with the objective to overcome the shortcomings of a layered approach that lacks of information sharing across protocol layers, forcing the network to operate in a suboptimal mode. For this reason, while for the sake of clarity in the following sections we present the challenges associated with underwater sensor networks following the traditional layered approach, we believe that the underwater environment particularly requires for cross-layer design solutions that allow a more efficient use of the scarce available resources.
However, although we advocate integrating functionalities to improve network performance and to avoid duplication of functions by means of cross-layer design, it is important to keep an eye on ease of design by following a modular design approach. This also allows improving and upgrading particular functionalities without the need to re-design the entire communication system.