Abstract:

Improving sync accuracy in Wireless Sensor Network (WSN) usually causes additional sync overhead. To optimize the compatibility of higher sync accuracy and lower sync overhead, a time sync algorithm based on the component decoupling fusion was proposed. With the two-way broadcast sync mechanism and the clock correlations, the clock deviations between synchronized nodes and reference nodes were estimated by the component decoupling fusion. Besides, the computation for the weights of different components was discussed according to the linear unbiased minimum variance estimation. The simulation results show that, without the additional energy consumption, the proposed algorithm can improve its sync accuracy after 20 sync rounds by 4.52μs, 13.8μs and 25.48μs, compared to PBS (Pairwise Broadcast Synchronization), TPSN (Timing-sync Protocol for Sensor Network) and RBS (Reference Broadcast Sync) respectively.

Key words: Wireless Sensor Network (WSN), time synch, component decoupling fusion, sync accuracy, sync overhead

摘要：

无线传感器网络(WSN)时间同步精度的提高通常以增加额外同步能耗为代价，针对时间同步的高精度与低能耗的权衡问题，提出一种基于分量解耦融合的时钟同步算法。该算法结合双向广播同步机制和节点间时钟的相关性，利用分量解耦融合思想计算被同步节点与基准节点间的时钟偏差，同时根据线性无偏最小方差准则估计各个分量加权参数的取值。仿真结果表明，该算法能够在不增加额外同步能耗的条件下，相比PBS、TPSN和RBS算法，20轮同步后同步精度分别提高4.52μs、13.8μs以及25.48μs。

关键词: 无线传感器网络, 时间同步, 分量解耦融合, 同步精度, 同步能耗