Study On Joint Estimation Methods Of Clock Skew And Offset Based On Timestamp-free Synchronization

Time synchronization is one of the considerable fundamental technologies for wireless sensor networks(WSNs),and its role is to provide a uniform time reference for system to ensure collaborative operations among nodes.At present,most existing time synchronization algorithms in WSNs require full timestamp information so as to estimate clock offset and skew to maintain clock synchronization among nodes.In recent years,timestamp-free synchronization mechanism that only uses limited timestamp information to achieve clock parameter estimation.It does not need to interact with any timestamp,and synchronization information is transmitted via receiver’s predefined response to sender.So parameter estimation function can be seamlessly embedded into the existing network data traffic significantly reducing the synchronization energy consumption.However,the existing timestamp-free synchronization scheme can only estimate clock skew,not clock offset.It becomes a challenge to estimate both clock skew and clock offset with limited timestamp information.To address the above problem,an intensive study on joint estimation methods of clock skew and offset based on timestamp-free synchronization are provided in this thesis.The main work of the dissertation includes:1.This thesis introduces the basic theory of time synchronization for wireless sensor networks,including clock model and message transmission delay,and make a detailed analysis of two timestamp-free synchronization methods and mechanisms of combining timestamp-free synchronization with implicit synchronization and one-way message transmission,respectively.2.Aiming at the issue that existing timestamp-free synchronization mechanisms can only estimate clock skew but not clock offset,a partial timestamp information synchronization protocol based on timestamp-free interaction is proposed.By presetting the local time of reference node returning the acknowledgement frame to transmit partial timestamp and response time information,node can jointly estimate clock skew and offset without exchanging the timestamp,and realize synchronization between nodes.Based on the partial timestamp information synchronization communication process,clock parameter estimation model of active is established.Considering Gaussian random delay,clock skew and offset are jointly estimated using maximum likelihood estimation(MLE)method,and the corresponding performance lower bounds are derived to measure the performance of the estimators.Considering exponential random delay,clock skew and offset are estimated simultaneously using the best linear unbiased estimation(BLUE)method.Finally,the effectiveness of the proposed estimation algorithms under two random delay models is verified by simulation,and the performance and computational complexity are compared with existing algorithms.3.For the problem that silent node can only estimate clock skew but not clock offset in energy-efficient scenario combining timestamp-free synchronization and implied synchronization,a joint estimation method of clock skew and offset for implied node based on timestamp-free interaction is proposed.Embedding the proposed partial timestamp information synchronization method into the implicit synchronization mechanism,implicit node can estimate clock skew and offset simultaneously by listening to the message interaction between reference node and active node,and achieve synchronization with reference node.In the process of synchronization,the implicit node does not need to obtain timestamps of any other node,but only needs to receive the synchronization message,which significantly reduces the synchronization energy cost.Based on the communication process of the partial timestamp information for three nodes,clock parameter estimation model of the implied node is established.For Gaussian random delay and exponential random delay,clock skew and offset are estimated using MLE and BLUE method,respectively.Finally,the performance of the proposed estimators of clock skew and offset for implicit nodes is verified via simulation,and the performance is also compared with the similar estimators.In this thesis,the joint estimation methods of clock skew and offset based on timestamp-free synchronization are investigated.A partial timestamp information synchronization protocol is proposed.For active node and implied node,by utilizing the statistical signal processing technology,the clock skew and offset under Gaussian random delay and exponential random delay are jointly estimated,respectively.Finally,simulation results show the effectiveness of the proposed joint estimation algorithms.This work has certain reference significance for the application of clock synchronization scheme with limited timestamp information in practical wireless sensor networks.