Low Energy, Multimodal Sensing for Wireless Underwater Monitoring Networks

Monitoring the activity in our oceans is critical for marine life conservation, underwater threat detection, subsea asset and pollution monitoring. However, this exercise is disrupted by the unpredictable nature and scale of the ocean as a transmission medium. Whilst acoustic sensing is the ocean's gold standard due to its attenuation benefits over light and radio waves, a multi sensor system capable of operating in a wide area subsea network would afford enhanced detection capabilities when compared to an isolated 'acoustic only' system. As the nodes will be isolated across a wide area in an ocean-based network, it is important to consider their power constraints. This is accompanied with cost consideration as the network may be made up of hundreds or thousands of sensor nodes.

This project aims to address these challenges by developing a multi-modal sensing unit which allows for reliable event detection using a combination of sound, magnetic field and temperature data. Combining three sensing modes affords the prospect of a customisable multi-use product capable of detecting a variety of underwater events. Using novel signal processing techniques and a strict low-power, low-cost design approach, this project aims to develop the sensing front-end for Newcastle University's developing NanoModem underwater communication network.

As with most low cost and low power engineered systems, the capabilities will struggle to compete with a state-of-the-art unconstrained counterpart. This is where the benefits of this project being part of a widely distributed network become apparent. Firstly, the geographical area that this system could be capable of covering is theoretically endless. With a current 2km communication range between nodes on the NanoModem v1.0, the potential for large scale communication is very achievable. Secondly, given an accurate positioning capability of each node, the post processing of multiple sensor's data could give an indication of target localisation. Finally, as one of the features of the NanoModem's communication network is the ability to multi-hop data between nodes, sensor results can be conveyed very quickly back to shore without the need for one centralised hub for each cluster of nodes.

The overall key project aims are to produce a system that will:
* detect signals of interest using a multiple sensor approach
* perform accurate classification on the successfully detected signal
* transmit compressed data through the NanoModem communication network
* maintain a low-power, low-cost development approach throughout the project
These project aims can be further categorised into objectives to show the specific research areas of interest.

Cetacean Acoustic Signatures - Research and analyse the unique acoustic signatures that are used by whales, dolphins and porpoises (cetaceans) for purposes such as communication and echolocation.

Underwater Threat Signatures - Investigate the variety of underwater threats which may fall in the region of this project's sensing capabilities (i.e. acoustic, magnetic and temperature).

Detection Methods - Develop and simulate detection algorithms using MATLAB to test against known target signals of interest. The objective of this section is to achieve the highest successful detection rate with the lowest computational overheads in detecting targets.

Classification Methods - Once a positive detection (signal of interest) has been identified the next step is to classify the signal to produce useful data for transmission.

Data Transmission - To protect the health of marine life and also maintain a covert status in the detection of underwater threats, it is important to sanction the transmission of data in a selective manner.

Power and Cost - At the heart of this project's aim is the practical application of a multimodal sensing unit which is capable of being produced on a large scale at low cost.