Research On Underwater Sensor Deployment Based On Grid Division And Virtual Forces

Comparing with the traditional wireless sensor networks, Underwater Sensor Networks have features including sparse node deployment, dynamic network, complex environment, acoustic communication, limited energy. Because of the expensive underwater sensor nodes, large-scale intensive deployment cannot be reached. So the random deployed underwater sensor networks often reach low coverage efficiency. Although the certain deployment can achieve high coverage efficiency, it cannot apply to the too large area or the unreachable area, such as harsh environments, difficult terrain area. The major works of this thesis are about three-dimensional underwater sensor deployment based on the characteristics of underwater sensor networks, combined with the research of traditional 2D sensor networks.The characteristics of underwater sensor networks and the structure of underwater sensor networks and underwater sensor nodes are firstly summarized. The technology of underwater sensor network deployment is overviewed. Besides, the three-dimensional polyhedron filling theory is introduced. And similarities between underwater node deployment and three-dimensional polyhedron filling theory are expounded. These studies provide preparation for the next step research of underwater sensor network deployment issues.Aiming at the characteristics of underwater sensor networks, combining with three-dimensional polyhedron filling theory and virtual potential field theory in wireless sensor networks, a deployment strategy based on grid division and virtual forces is proposed. In this strategy, the problem of three-dimensional underwater sensor network deployment is reduced to the problem of two-dimensional pre-deployment. It completes the grids division of sensor nodes on the surface of water, uses a mature model of two-dimensional sensor network deployment, makes introduction of virtual force algorithm in the traditional two-dimensional sensor networks. Then, it generates body centered cubic lattice underwater surveillance network whose Voronoi cells are truncated octahedron by controlling the moving in the vertical direction of nodes. Simulation result shows that in this deployment strategy the network can achieve full coverage with a smaller number of nodes. It can effectively reduce the cost of deployment of the network.Aiming at the problems including limited energy, complex environment, more node failures and so on in underwater sensor networks, a coverage detection mechanism based on cubic grid division is proposed. This mechanism makes rational use of geometry, expands the two-dimensional grid division theory to 3D applications. The coverage of entire network is detected by generating virtual three-dimensional cubic grids, monitoring whether each cell is covered by at least one sensor node. If all the cells are covered, the entire target area is covered. And introduces a parameter v defined to the cover matrix of the network, through which the size and location information of holes in the entire network can be effectively known. Simulation result shows that the mechanism can effectively detect coverage holes in the entire network with less relative error. It is suitable for large-scale applications of underwater sensor networks.