Study On Parameter Estimation Methods For Clock Synchronization With Timestamp Free Exchanges In Wireless Sensor Networks

Time synchronization is one of the key technologies to maintain the stable operation of wireless sensor networks(WSNs),which ensures a common timescale for all the nodes in WSNs and plays an important supporting role in the basic applications of WSNs,such as data fusion,power management,sleep scheduling,etc.As a novel synchronization mechanism,timestamp free synchronization achieves parameter estimation for clock synchronization between nodes just by sending and receiving common data packets according to certain rules,which does not need to exchange timestamps or construct dedicated synchronization messages.Compared with most of the current synchronization mechanisms based on timing message exchanges,it has the advantages of energy-efficiency,security and easy-integration,thus attracting much attention in recent years.Despite the excellent performance of the timestamp free synchronization mechanism,the research on the joint optimization of it with implicit synchronization mechanism or one-way message dissemination mechanism is still in the blank,and there is also a lack of relevant research under time-varying clock skew model,which restricts the application of timestamp free synchronization in actual networks.In view of the above problems,this dissertation makes an intensive study of parameter estimation methods for clock synchronization with timestamp free exchanges in WSNs.The main contributions of this dissertation include:1.Considering the superiority that the silent nodes within the overlapping communication range of pairwise synchronization nodes can achieve synchronization only by listening the synchronization procedures between this pair of nodes,the two low-power synchronization mechanisms,e.g.,implicit synchronization and timestamp free synchronization,are combined for joint optimization,and a clock skew estimation method for silent node with timestamp free exchanges is proposed,which can further reduce energy consumption for clock synchronization.According to the established timestamp free parameter estimation model,the maximum likelihood estimator(MLE)and corresponding Cramer-Rao Lower Bound(CRLB)for clock skew of silent node are derived.Meanwhile,another simplified estimator is also derived by eliminating the nuisance parameters.Simulation results show that the proposed MLE and simplified estimator are both effective.2.To solve the problem that the timestamp free synchronization protocols based on linear clock model cannot provide clock offset estimation,a hybrid synchronization method based on timestamp free synchronization and one-way message dissemination is proposed to jointly estimate the clock skew,clock offset and fixed delay.Since the timestamp free synchronization can be seamlessly embedded in network traffic and the one-way message dissemination can be simply implemented in parallel to broadcasting mechanism,the hybrid synchronization method can be easily integrated into practical WSNs.After analyzing the hybrid synchronization process and establishing the synchronization model,the MLE for joint estimation of the clock synchronization parameters is derived,as well as corresponding CRLB.The performance advantages of the hybrid synchronization method are proved by simulations.3.Aiming at the problem that most of existing researches related to timestamp free synchronization assume that the clock skew remains constant over a period of time,a clock skew tracking algorithm based on Kalman filtering with timestamp free exchanges is proposed under the more general time-varying clock skew model.Based on the timestamp free synchronization protocol,the observation model is established,and the first-order autoregressive model is applied to reflect the time-varying characteristic of the clock skew,and then the Kalman filter is used to achieve the long-term tracking of clock skew.Simulation results show that the proposed algorithm is able to accurately track the real skew.In this dissertation,the timestamp free synchronization mechanism is combined with the implicit synchronization mechanism and the one-way message dissemination mechanism for joint optimization,respectively,and the research under the time-varying clock skew model is also conducted.From the perspective of statistical signal processing,the clock synchronization parameters are estimated effectively,and the efficiency of the proposed algorithms are validated by simulations,which provides a certain referential significance for the application of timestamp free synchronization in practical networks.