IEEE 1588 Based Time Synchronization Method For Industrial Wireless Sensor Networks

Time synchronization is the critical technology to realize the application of industrial wireless sensor network,such as data fusion,time slot scheduling,and multi-machine cooperation.IEEE 1588 Precise Time Protocol(PTP)is based on UDP/IP protocol,which is easy to operate,low bandwidth requirement,portability,etc.,These features provide an opportunity for IEEE 1588 protocol’s application in industrial wireless sensor networks.But the inherent randomness of sending and receiving packets in wireless networks,the uncertainty of industrial site environmental noise,and the asymmetry of uplink and downlink,etc.,constraint the improvement of IEEE 1588 time synchronization protocol performance in industrial wireless network environment.Therefore,this paper constructs a precise clock model,and then designs related optimization algorithms to improves the accuracy and reliability of the IEEE 1588 time synchronization protocol for industrial wireless sensor networks.The main research contents are as follows:(1)Concerning the dynamic changes of link delay,clock timing interference,and uncertain timestamp acquisition caused by complex link environments and temperature fluctuations in industrial wireless sensor networks.The feasibility of applying PTP protocol in industrial wireless sensor network is demonstrated firstly.The limitations are further analyzed by integrating the clock timing interference and asymmetric link delay noise in the PTP two-way time synchronization process.Then,the dynamic model is used to establish the clock state space model and observation mode to accurately describe the clock evolution and time synchronization mechanism.Finally,based on the Matlab/Simulink simulation environment,the PTP time synchronization simulation platform is designed.(2)In view of the lack of specific clock offset and skew estimation methods in PTP protocol,and the inability of existing filtering algorithms to adapt to dynamic noise interference,a reverse adaptive Kalman filter algorithm is studied.The rationality of the noise statistical model is evaluated by using the clock state normalized innovation ratio of the reverse estimation and the forward estimation.Subsequently,the process noise of the clock state is dynamically adjusted by the preset detection threshold to achieve high-precision estimation of clock parameters.This algorithm can solve the problem of Kalman filter divergence caused by uncertain noise.The simulation results show that the algorithm can effectively improves the accuracy and reliability of PTP time synchronization under the uncertain noise model.(3)In industrial wireless sensor networks,the two-stage filtering algorithm is designed for the problem of anomalous asymmetric delay in message transmission caused by high load.The algorithm includes two parts: outlier calibration and clock parameter estimation.Firstly,the asymmetric delay ratio R is calculated based on PTP message interactions in a high-load network environment,and outliers are filtered using R.Secondly,the clock information at the moment of outlier generation is optimally estimated by Kalman filter to ensure that accurate clock synchronization parameters are available to adjust the local clock.The simulation results show that the two-stage filtering algorithm exhibits high accuracy and stability in the situation with link asymmetry.