Research On Deployment Policy Of Underwater Wireless Sensor Networks

As the expansion of Wireless Sensor Networks, Underwater Wireless SensorNetworks(UWSNs) has aroused extensive attention. By deploying wireless sensor nodes has thefeature of acoustic communication into the given water area, UWSNs can be applied in many areaslike marine environmental monitoring, geological calamity forewarning, exploitation of marineresources, navy affairs and so on. Broad application prospects arouse the world of science’s interestand enthusiasm. However, UWSNs’natural defects such as low available bandwidth, highly varyingmultipath and large propagation delays, couple with objective factors like the complexity of theunderwater environment, the uncertainty of the monitor targets and the influence of the waterflowbring about new challenges in nodes deployment issue.Thus, the whole thesis focus on the node deployment issue. After the contrast and analysis ofdifferent node deployment policies, a Voronoi-based Depth-Adjustment algorithm for UWSNs wasproposed. The algorithm use Voronoi polygon area as the measure of nodes density. With the help ofself-depth-adjustment scheme, nodes that deployed randomly in the water surface would divide thetarget region into different layers, therefore, the underwater coverage was extended, and the nodedeployment process was simplified. Besides, experiments turned out that this approach have a goodperformance in network coverage and connectivity. However, the experiments also pointed out thatconnectivity of the network is very sensitive to RT/RS.At the end of the paper, a temperature data acquisition system which was based on GAINZ—aWSNs development kits developed by Institute of Computing Technology, Chinese Academy ofSciences—was designed. By expounding the realization processes of different functional modules,we add the core theory of the Voronoi-based Depth-Adjustment algorithm into the system. Thesystem tests not only indicates the system’s functional completeness and the algorithm’s feasibility,but also verify that in some cases the algorithm capable of optimize the node number.