基于虚拟力驱动的传感器网络节点定位

为了解决传感器网络节点三边测量定位的全局优化问题,提出了一种基于虚拟力驱动的节点最优相对定位方法。该方法建立了满足节点位置最优最小二乘估计需要的虚拟势能场模型,给出了一种基于虚拟势能场梯度的虚拟力定义新方法,证明了利用该虚拟力驱动节点定位的收敛性,提供了保证节点定位结果全局最优的方法。仿真分析表明,基于虚拟力驱动的节点相对定位方法能够解决节点相对位置的最优估计问题。

Location Method for Sensor Networks Nodes Based on Virtual Force Driving

In order to get the globally optimal solution of trilateration location for sensor network nodes, a novel virtual force based optimal relative location method is proposed. The new method first constructs the virtual field potential model, which meets the needs for optimal position estimation of nodes in sense of least mean square sense. Then a new definition for virtual force is presented based on the gradients of virtual field potential. The convergence of the node location method driven by the virtual force is proved. The technique that produces the globally optimal location is provided. Simulation results shows that the virtual force based location method can get the optimal estimations of the relative positions of sensor nodes.