Font Size: a A A

Clock Parameters Tracking Based On Timestamp-Free Mechanism In Wireless Sensor Networks

Posted on:2023-08-18Degree:MasterType:Thesis
Country:ChinaCandidate:R LuFull Text:PDF
GTID:2568307031988669Subject:Control engineering
Abstract/Summary:PDF Full Text Request
For WSNs,clock synchronization is a key technique to ensure a unified time reference,and it is a significant basis for the coordination of all devices in the network.In addition,the basic functions of WSN such as data fusion and resource scheduling require the support of high-precision clock synchronization.Timestamp-free synchronization is an energy-efficient synchronization method proposed in recent years,and its characteristic is that there is no need to transmit any timestamp information and construct specific synchronization packets during the synchronization process.Therefore,this method can be easily embedded into the sending and receiving process of general data packets,so that the dataflow can be used to achieve synchronization,which significantly reduces the synchronization energy consumption and effectively avoids potential attacks on timestamp.Most of the existing research on timestamp-free synchronization assumes that clock parameters remain unchanged in a period of time,but due to the influence of environmental factors,the clock crystal oscillator is actually unstable,resulting in dynamic changes in clock parameters.In view of the problems mentioned above,this dissertation takes the time-varying clock model into consideration,and conducts an intensive study on clock parameters tracking based on timestamp-free mechanism in WSNs.The main work of this dissertation includes:1.Considering the sensitivity of devices to energy consumption in the application of practical WSN and the instability of the clock crystal oscillator during long-term operation,the timestamp-free synchronization mechanism is extended to the scenario of time-varying clock parameter.A clock parameters tracking model based on Gaussian Markov process is established,and a clock skew and offset joint tracking algorithm of active node for timestamp-free synchronization based on Extended Kalman filter(EKF)is proposed,which makes up for the defect that the existing timestamp-free synchronization can only track clock skew.The proposed algorithm achieves long-term dynamic tracking of clock skew and offset while generating almost no additional synchronization energy consumption.Meanwhile,corresponding Posterior Cramer-Rao Lower Bound is derived to evaluate the performance of the proposed tracking algorithm.Finally,the advantages of the proposed tracking algorithm are demonstrated by the simulation and analysis.2.Taking advantage of the feature that the silent node in receiver-only synchronization(ROS)can achieve clock synchronization only by overhearing paired messages,the clock parameter tracking of timestamp-free synchronization and implicit synchronization are jointly designed.A clock parameter tracking model for silent nodes is established,and the EKF is used to present a clock skew tracking algorithm of silent node for timestamp-free synchronization.Then,corresponding Posterior Cramer-Rao Lower Bound is derived.According to the characteristics of timestamp-free synchronization and the silent node,the proposed algorithm can realize long-term tracking of clock skew without requiring dedicated command frames for clock synchronization and additional packet exchange,further reducing energy consumption and communication overhead.The simulation results show the effectiveness of the proposed algorithm.3.Aiming at the problem of insufficient tracking accuracy and robustness of clock skew for silent nodes based on timestamp-free synchronization,this dissertation considers the characteristics of silent nodes listening to paired synchronization messages from multiple active nodes,and the Information Fusion Estimation Theory and Multi-sensor Information Fusion Technique are applied to clock skew tracking of silent nodes for the first time.A synchronization model of the silent node based on multiple observations from multiple communication paths is established,and a multi-observation fusion tracking algorithm of silent node without timestamp interaction is proposed.The simulation results and analysis show that the clock parameters tracking accuracy and robustness of the silent node are significantly improved after implementing the multi-observation fusion tracking algorithm.Based on time-varying clock parameters which are more practical,this dissertation adopts EKF and Multi-sensor Information Fusion Technique to conduct clock parameters tracking research for various scenarios of timestamp-free synchronization.Meanwhile,simulation results show the effectiveness of the proposed algorithms,which provides a certain theoretical support for the practical application of timestamp-free synchronization.
Keywords/Search Tags:time synchronization, wireless sensor networks, timestamp-free synchronization, parameter tracking, Kalman filter
PDF Full Text Request
Related items