无线传感器网络分布式同步协议

Distributed synchronization protocol for wireless sensor networks

针对无线传感器网络同步问题,提出分布式时间同步和分布式数据同步的解决方法.前者要求簇头网络进行局部信息交互,并采用低通滤波技术去除噪声干扰;后者为节点提供网络数据均值,要求簇头网络执行比例一致性算法,簇头在迭代过程中引入簇内节点数量.实验结果表明,分布式时间同步具备抗噪声能力,该算法在前期收敛速度最快.网格状网络和随机网络实验表明,分布式时间同步和分布式数据同步的通信开销非常低,它们的收敛速度均高于普通数据同步.

Distributed time synchronization and distributed data synchronization are proposed for the synchronization problem in wireless sensor networks. The former requires the duster head network to execute local information exchange, and it adopts low-pass filtering to remove the noise interference. The latter provides network-wide data mean to nodes, and it requires the cluster head network to execute the proportion consistency algorithm, in which the number of nodes within a cluster is introduced during the iterative process. Experimental results show that the distributed time synchronization maintains anti-noise performance, and that the algorithm converges fast in the earlier stage. Grid-like network and random network experiments show that distributed time synchronization and distributed data synchronization have a low communication overhead, and their convergence rates are faster than that of general data synchronization.