Research On High Precision Wireless Time Synchronization Technology For Mobile Platform

In recent years,with the growing demand for unmanned systems in various fields,a great deal of motion platforms such as unmanned aerial vehicles and unmanned vehicles have emerged.They have been widely used in industry,agriculture,service industry and even national defense.In practical applications,time synchronization is the prerequisite and guarantee for collaborative work of motion platform nodes.Therefore,the research on high-precision time parameter estimation of moving platform nodes has attracted more and more attention.Based on the perspective of statistical signal processing,thesis studies the time synchronization technology from two aspects: stationary platform nodes and motion platform nodes.The main research work of thesis is as follows:(1)Aiming at the static platform node,thesis proposed a novel time synchronization algorithm with unknown deterministic propagation delay.Firstly,thesis analyzed the Sender-Receiver Synchronization(SRS),SRS didn’t make effective use of broadcast information of nodes.Then taking advantage of the broadcast characteristics of the signal and Receiver-Receiver Synchronization(RRS),a time synchronization algorithm integrating SRS-RRS is proposed.And thesis derived the Cramér-Rao Lower Bound(CRLB)corresponding to SRS algorithm and SRS-RRS algorithm with or without the consideration of clock frequency offset.Thesis gave the closed-form solutions of the two algorithms by Maximum Likelihood Estimate.Simulation results showed that SRS-RRS algorithm has higher synchronization accuracy than SRS algorithm and can achieve CRLB,and its accuracy will be improved with the increase of the number of nodes which will be synchronized.(2)Aiming at the motion platform node,thesis constructs a time synchronization system model and proposes three time-synchronization algorithms.Without considering the clock frequency deviation and the lack of measuring means for the speed and position of the motion platform node,thesis modeles the time-varying distance caused by node motion.The system model of the time synchronization of the motion platform is established with the help of the anchor node which has been synchronized and the position is known,and thesis derives the CRLB of the system.Thesis proposes Two Step Weighted Least Squares(TSWLS)algorithm and Semi-Definite Relaxation(SDR)algorithm based on Least Squares(LS).Considering that MLE is asymptotically unbiased and can asymptotically reach the CRLB,thesis proposes a Newton iteration method based on MLE.In order to prevent its divergence,the solutions of TSWLS and SDR are used as the initial solutions of Newton iteration method.The simulation analysis shows that both TSWLS-iteration method and SDR-iteration method can achieve the corresponding CRLB of the system,but the solution complexity of TSWLS-iteration method is lower,and both accuracy and computational complexity can be made balanced.(3)Based on the system model discussed above,thesis constructs a system model with anchor node error and proposed three time-synchronization algorithms.Considering the accuracy of the location estimation algorithm and clock parameter estimation algorithm in the actual environment,thesis adds the anchor node position error and the clock phase bias error into the above model,and then derives the corresponding CRLB and studies the relation and difference with the system CRLB without considering the anchor node error.Also based on this system model,three methods are proposed: TSWLS method based on LS problem,SDR method based on LS problem and iteration method based on MLE problem.The simulation results show that SDR method can provide better initial solution than TSWLS method,and SDR-iteration method has the estimation accuracy close to CRLB.