随机网络中分布式节点高精度自定位算法

Randomly Distributed Node in the Network High Precision since the Localization Algorithm

随机网络不确定性高,较为复杂,当前节点定位算法无法准确对随机网络中分布式节点进行准确定位,且适应性差。为此,提出一种新的基于标记传递的随机网络中分布式节点高精度自定位算法,标记传递算法将随机网络中的分布式节点用有标记数据和无标记数据进行描述,依据节点间的相似度将节点标记传递至其相邻节点。对节点定位问题进行描述,介绍了标记传递算法,在此基础上通过标记传递算法对随机网络中分布式节点进行高精度自定位,定位过程主要分为离线训练阶段与在线自定位阶段,给出定位详细实现过程,对定位结果进行滤波处理。实验结果表明,所用所提算法对随机网络中分布式节点进行自定位,定位精度高,适应性强。

Because of the high uncertainty and complexity of the random network,the current node localization algorithm can not accurately locate the distributed nodes in the random network. To this end,a novel distributed random network token node in high precision self localization algorithm was proposed based on token algorithm distributed nodes in random network with labeled data and unlabeled data is described,based on the similarity between nodes will be transmitted to the adjacent node labeled nodes. To describe the problem of node localization,the marker transfer algorithm based on the marker passing algorithm of high precision self positioning of distributed random nodes in the network,the positioning process is divided into off-line training phase and the online self positioning stage,detailed implementation process were given,the result of locating filtering. The experimental results show that the proposed algorithm has the advantages of high localization accuracy and strong adaptability to the distributed nodes in the random network.