Research On Time Synchronization Algorithm In Wireless Sensor Networks

Wireless sensor networks are widely used in the fields of military,industrial monitoring,environmental detection,smart life and so on due to their strength of low cost,strong scalability and high stability.Time synchronization is a prerequisite for correct data detection,transmission and fusion in wireless sensor networks.Due to the constraints of computing capacity,storage capacity,communication range and energy of sensor nodes,there are still many difficulties in the development of time synchronization algorithm in wireless sensor networks.The consensus time synchronization algorithm of wireless sensor networks is studied in this thesis.The main works are as follows:First,a new performance target of time synchronization algorithm,called long-term stability in the thesis,is studied for wireless sensor networks that need work for more than half a year in addition to synchronization accuracy,convergence rate and energy consumption.Long-term stability describes the relationship between the synchronization accuracy and the start time of the algorithm in wireless sensor networks which need to synchronize its time several times.Long-term stability reflects whether a time synchronization algorithm for wireless sensor network can keep synchronization accuracy stable long time later,and whether it can make a wireless sensor network keep time synchronization in its whole life.Second,because of the problem that the classical consensus time synchronization algorithm does not have good long-term stability,and the transmission delay and packet loss are not estimated sufficiently,the thesis proposes an average consensus time synchronization algorithm with long-term stability.The algorithm uses physical layer timestamp to reduce the transmission delay.After that,the transmission delay is estimated and compensated.In addition,according to the occurrence of packet loss,the algorithm adjusts the parameters in the iterative process to reduce the influence of packet loss on the convergence rate.In this algorithm,the estimation of clock skew and clock offset are carried out separately to avoid the large fluctuation in clock offset estimating caused by the compensation of clock skew when hardware time is very large in the wireless sensor network which need work for more than half a year,which improves the long-term stability of the algorithm.Thirdly,the synchronization accuracy and long-term stability of the proposed algorithm are proved by simulations and experiments.Under various simulation and experimental conditions,the synchronization accuracy of the proposed algorithm can reach less than 3?s,and even less than 1?s under some of conditions,which is obviously higher than the traditional consensus time synchronization algorithm.When the time synchronization runs at 1 week,1 month or 1 year after the wireless sensor network starts,the proposed algorithm can still maintain the same synchronization accuracy,i.e.it has good long-term stability.