摘要

Two-way time-delay exchange algorithm represented by the precision time protocol (PTP) is one of the fundamental clock synchronization approaches in wireless sensor networks. However, the accuracy of the algorithm is declined due to the random asymmetric delay in wireless links. Although statistical estimation techniques can improve the synchronization accuracy, they need to collect large samples. This will lead to slow convergence and cause memory problems for resource-constrained devices in wireless sensor networks. A synchronization algorithm is proposed to realize fast clock synchronization which is robust to the unknown delay distributions based on grey prediction using the GM(1,1) model.