The Experiment Research On Sensing Area Model Of Node In Wireless Sensor Network

The knowledge of a sensor'sensing area is the foundation of wireless sensor network (WSN) design and applications. Simulations of most of the current research on WSNs assume an ideal disk model of the sensing area, which, as has been shown however, can lead to discrepancy between simulation performance of some algorithms, such as the algorithms of MAC, routing, localization, covering, and topology control, and practical performance. Therefore, the building of more realistic sensing area models is crucial.In this thesis, a wireless sensor network experiment platform was firstly built up and the differences between sensor nodes were calibrated, which laid the foundation for the later study. By using time slot assignment, problem of heavily packet loss was solved. With the routing protocol, all the nodes in the network can send data to the root. Then, based on the analysis of traditional uni-directional wireless propagation model, the effect of node heights, send power and flatness of the ground on wireless propagation model was studied in three linear topological networks. Then a piecewise-fitting model and a logarithmic-attenuation-based error-compensation model were built. Furthermore, based on the analysis of the omni-directional propagation, four disk models with different radius were conducted, to study the direction of power loss and the correlation between neighboring nodes, to compare the sensing areas in different directions, and to analyze the impact of node positions on the sensing area, and a more real sensing area model RSAM and an improved RRSAM model with fluctuant change in different directions were then build up.Three linear topology experiments'results show that uni-directional propagation model sorts into linear region and fluctuant regional, and the size of the linear area is affected by the node height and flatness of the ground; on the comparison of the ranging error between piecewise-fitting model and the logarithmic-attenuation-based error-compensation model, results show that the ranging performance of the logarithmic-attenuation-based error-compensation model is better than that of piecewise-fitting model. Through four circular topology experiments with different radius, the results show that with the same radius, the received power of points in different directions are different but relevant; and studies also have shown that RSAM can reflect the macroscopic fluctuation of actual node perception regional, while RRSAM can reflect not only the macroscopic fluctuation but also the microscopic fluctuation of the actual node perception regional, which have great significance on the design and application of wireless sensor network.