欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
提出一种动态组簇的协同定位方法，用于基于传感器网络的目标定位和跟踪．该方法包括数据融合算法和虚拟簇漂移（virtual cluster shift，VCS）机制两部分．数据融合算法部分采用均值漂移（mean shift）算法．虚拟簇漂移机制分布式地在组织目标周围的锚节点建立临时簇．簇首管理簇成员，收集感知数据，执行融合算法．当虚拟簇无法锁定目标时，簇首指定离目标最近的簇成员担任新簇首，簇的成员也进行更替，由此将虚拟簇移动（shift）到合适的位置．分析和仿真结果显示，采用动态组簇的协同定位方法跟踪目标可以大幅度降低通信开销，产生的通信量仅为以往集中式定位算法开销的1／3．
英文摘要：
A new localization method is proposed for wireless sensor networks. The method uses iterative localization algorithm and distributed clustering scheme to decrease generated communication packets and increase the locating accuracy. The clustering scheme, named virtual cluster shift （VCS）, makes self-organized sensor nodes of a wireless sensor network dynamically build one or more virtual clusters to localize and track mobile targets. When one or more sensors detect a target, a sensor node is selected to be the cluster head. The cluster head is in charge of organizing neighboring sensors to form a virtual cluster, fusing data and estimating the target＇s states such as location, velocity of mobile target. When the target is moving out of the monitor area of current virtual cluster, the cluster head gives place to one of his cluster member who is the nearest to the target. The new cluster head forms his virtual cluster by replacing some of the old cluster members. A mean shift based algorithm is used to iteratively estimating the target＇s position using data gathered inside virtual cluster. An optimal weight of the kernel function is given to minimize the variance of estimation. Analysis and simulation results show that the communication overhead and jitter of the proposed algorithm is less than 1/3 of the centralized algorithm.
位置：
同期刊论文项目
