Abstract:As one of the support technology of wireless sensor network applications,node localization has been paid much attention to both academia and industry all the times.Existing localization methods are not suitable for 3-dimensional(3D)terrains due to their design for planar applications.According to the shortage of some existing algorithms proposed for 3D,in this paper we present a dynamic localization scheme based on spherical coordinates which translates the localization problem into a solution of multiple linear equations and utilizes the Cramer′s rule to eliminate the probability of multiple solutions or void solution for the equations.Simulation results indicate that the relative localization error is 50.7 percent when 100 sensor nodes are deployed in 100*100*100m3 scenarios with 10 meters′ communication range and the anchor node density is 10 percent.The least square estimate(LSE)is applied to our algorithm to estimate the sensor nodes′ positions and a result filtration strategy is used to reduce the localization error,the experiment data show that the improved solution outperforms the original scheme by 32.2 percent.

A Localization Scheme Based on Sphere for Wireless Sensor Network in 3D