针对无线移动传感器网络(WMSN)自身具有移动性和分布式的特点,在现有分布式粒子滤波DPF的基础上,提出了一种实现分布式数据融合的分布式目标跟踪算法WDOT.该算法使用高斯混合模型和平均一致性滤波器实现了节点的分布式数据融合,提高了算法的稳定性,降低了通信开销.仿真实验通过与已有目标跟踪算法在通信开销、估计均方根误差等方面的对比,验证了WDOT算法的目标跟踪性能.%Aim at the feature of mobility and distributed of WMSN, this paper provided a distributed objects-tracking algorithm called WDOT (distributed objects tracking for WMSN). It was based on existing DPF and realized distributed data fusion. The algorithm put the Gauss mixed model and average consensus filter into local DPF of every node, and it could improve the stability of the algorithm with low communication cost. Simulating experiment compared some typical parameters with existing objects tracking algorithms, such as communication cost and MSE. The results show that WDOT algorithm has good objects tracking function for WMSN.
展开▼
