Abstract:

Concerning the dynamic change of link delay， clock timing interference， and uncertainty of timestamp acquisition caused by complex link environment and temperature fluctuation in Industrial Wireless Sensor Networks （IWSNs）， a time synchronization method based on Precision Time Protocol （PTP） in IWSNs was proposed. Firstly， the clock state space model and observation model were constructed by integrating the clock timing interference and asymmetric link delay noise in PTP bidirectional time synchronization process. Secondly， a reverse adaptive Kalman filter algorithm was constructed to remove the noise interference. Thirdly， the rationality of the noise statistical model was evaluated by using the clock state normalized innovation ratio of the reverse estimation and the forward estimation. Finally， the process noise of the clock state was dynamically adjusted after setting the detection threshold， thereby achieving precise estimation of clock parameters. Simulation results show that compared with the classical Kalman filter algorithm and PTP protocol， the proposed algorithm has the clock offset and skew estimation with smaller and more stable error standard deviations under different clock timing precision. The reverse adaptive Kalman filter can effectively solve the problem of Kalman filter divergence caused by reasons such as noise uncertainty， and improve the reliability of time synchronization.

Key words: Industrial Wireless Sensor Network (IWSN), time synchronization, Precision Time Protocol (PTP), clock state process noise, reverse adaptive Kalman filter algorithm

摘要：

针对工业无线传感器网络（IWSNs）中复杂链路环境、温度波动等造成的链路时延动态变化、时钟计时干扰、时间戳获取不确定等问题，提出一种基于精确时间协议（PTP）的IWSNs时间同步方法。首先，融合PTP双向时间同步过程的时钟计时干扰、非对称链路时延噪声，建立时钟状态空间模型和观测模型；其次，构建反向自适应卡尔曼滤波算法以滤除噪声干扰；然后，利用反向估计和正向估计的时钟状态归一化新息比值来评估噪声统计模型的合理性；最后，在设定检测阈值后，动态调整时钟状态过程噪声，以实现时钟参数的精确估计。仿真结果表明，相较于经典卡尔曼滤波算法和PTP，在不同时钟计时精度下，反向自适应卡尔曼滤波算法估计的时钟偏移和偏移率均有较小且更稳定的误差标准差，有效解决了噪声不确定等原因造成的卡尔曼滤波发散问题，提高了时间同步的可靠性。

关键词: 工业无线传感器网络, 时间同步, 精确时间协议, 时钟状态过程噪声, 反向自适应卡尔曼滤波算法