Research On Energy-saving Data Acquisition Techniques In Wireless Sensor Networks

As a rising technology of data acquisition, wireless sensor networks have a deep impact on people's work and life, which has attracted wide attention from domestic and foreign researchers. Data acquisition is a key component to realize the monitoring and sensing function of wireless sensor networks. The realistic contradiction on the limited energy of wireless sensor networks is serious and obvious. Rational selection of working nodes makes data acquisition of sensor networks more comprehensive. Optimization of area coverage can reduce data redundancy of sensor networks. So the node selection method and area coverage strategy affect the sensing results of monitoring area and can save much energy of sensor networks. Therefore, it is significative and valuable to research on data acquisition techniques of wireless sensor networks.In this thesis, the research on energy-saving data acquisition techniques in wireless sensor networks is mainly divided into two aspects. Firstly, for the problem that great redundancy and correlation in sensing data of multimedia sensor networks, a novel node spatial correlation model is designed and a node selection algorithm is proposed based on the correlation coefficients. The node selection algorithm focuses on reducing the spatial redundancy of sensing data and saving energy of nodes based on the spatial correlation model which describes the relevant characteristics between the observed images of nodes. Besides, we study the coverage problem in directional sensor networks for 3D complex terrains, and design new surface coverage algorithms. Based on a directional sensing model of nodes, these algorithms employ the ideas of grid division, simulated annealing, and local collaborative optimum to improve the area coverage ratio by optimizing the position coordinates and the deviation angles of the nodes, which results in the coverage enhancement for 3D complex terrains. We also conduct extensive simulations to evaluate the performance and effectiveness of our algorithms.