基于灰预测的PTP协议改进方法

发布时间：2018-03-14 17:45

本文选题：精确时钟同步协议　切入点：时钟同步　出处：《系统工程与电子技术》2017年08期 　论文类型：期刊论文

【摘要】：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。
[Abstract]:With the increasing demand of wireless network node clock synchronization and the increase of real-time application, the status of clock synchronization is becoming more and more prominent. In improving the user experience, improving the utilization rate of node resources, improving the stability of the system, etc. Clock synchronization is also of great significance. Precise time protocol PTPs are widely used in the clock synchronization mechanism. However, due to the existence of random asymmetric delay in wireless networks, The accuracy of PTP is reduced. Although the synchronization accuracy can be improved by statistical and estimation methods, a large number of samples are collected, which results in a slow convergence rate. For device nodes with limited resources, the problem of memory occupation is also caused. Firstly, the characteristic analysis of clock deviation between nodes in wireless network is given, and then a fast clock synchronization method based on the grey prediction model is proposed. The method is based on the grey prediction theory, which is aimed at the uncertain problem of inexperience and little data. For the clock synchronization of wireless network nodes with limited resources, the asymmetric delay is uncertain and the samples are few. The experimental results show that the proposed solution can achieve high precision and fast clock synchronization.
【作者单位】：西北工业大学自动化学院;
【基金】：西北工业大学基础研究基金(JC20110045);西北工业大学研究生创意创新种子基金(Z2016137)资助课题
【分类号】：TN915.04

【参考文献】