基于灰预测的PTP协议改进方法

Improved method of PTP protocol based on grey prediction

随着无线网络节点时钟同步需求的日益增加和实时性应用的增多,时钟同步的地位越来越突出,在改善用户体验、提高节点资源的使用率、提高系统稳定性等方面,时钟同步也有着重要意义。精确时钟同步协议(precision time protocol,PTP)广泛运用在时钟同步机制,然而,由于无线网络中存在随机不对称时延,使得PTP的精确度下降。尽管通过统计和估计的方法可以提高同步精度,但需要收集大量的样本,这将导致收敛速度减慢,对于资源有限的设备节点还会引发内存占用问题。首先给出了无线网络中节点之间时钟偏差的特性分析,然后提出了一种基于GM(1,1)的灰预测模型的快速时钟同步方法,该方法基于灰预测理论,灰理论是针对既无经验,数据又少的不确定问题。而对于资源有限的无线网络节点的时钟同步,不对称时延是不确定性的,同时样本很少,非常适合用灰理论进行时延估计。实验结果表明所提解决方案可以实现高精度的快速时钟同步。

As the increase of real-time application, the time synchronization between nodes in the wireless networks becomes more important. Time synchronization can help improving the user experience and raising the efficiency of resources and the stability of the systems. The precision time protocol （PTP） is widely used in clock synchronization mechanisms. However, the accuracy of PTP is declined due to the random asymmetric de-lay in wireless networks. Although statistical estimation techniques can improve the synchronization accuracy, large samples are needed to collect. This will lead to slow convergence, and it will cause memory problems for the limited resources device. Firstly, the property of clock offset is analyzed. Then, a synchronization algorithm based on grey prediction using the GM （1,1） model is proposed to realize fast clock synchronization. Grey theo-ry is for inexperienced, less data and uncertain problems. It＇s suitable for clock synchronization of wireless net-works. The experiment result shows that the proposed solution can obtain a fast clock synchronization with high-precision.