Clock Synchronization Technology Research Based On Improved ATS Protocol

In recent years,wireless sensor networks(WSN)are widely applied in various fields.Clock synchronization technology is an important component in the wireless networks,and it has been a hot and challenging research topic.Recently,consensus-based time synchronization protocols have attracted widespread concern and have been one of the hottest issues.Such protocols can not only has good robustness and scalability,but also can compensate both clock skew and offset.Unfortunately,there are two problems in the existing research works in terms with average consensus time synchronization(ATS)protocols.On aspect is slow convergence speed;the other is the low synchronization precision.In this paper,these problems are studied.Firstly,Wireless Sensor Networks is introduced in this thesis.The performance indexes of synchronization mechanism and the key factors affecting synchronization are also introducted,eapecially on synchronization principle and implementation method in centralized and distributed wireless sensor networks.Secondly,improved Bootstrap statistical algorithm,and under the condition of IEEE1588 synchronization protocol,using Kalman filtering slgorithm and improved Bootstrap algorithm to maintain the crystal oscillator stability,respectively.It is verified by MATLAB simulation that improved Bootstrap method is superior to the traditional Bootstrap method in the aspect of the calibration of frequency offset and time offset,and improved Bootstrap filtering algorithm can realize crystal oscillator stability fastly.Thirdly,use adaptive weighted algorithm and particle swarm optimization algorithm respectively to set up a data fusion model,which the optimal fused estimation is achieved.It is verified by MATLAB simulation that the model of particle swarm optimization algorithm based on natural selection can improve the accuracy of data fusion.At the same time,improved Bootstrap method combined with the particle swarm optimization fusion algorithm,then,the particle swarm optimization algorithm based on natural selection can effectively fuse the network nodes under the small sample condition,and originates a virtual clock with high precision.Finally,this paper studies the convergence speed and synchronization accuracy of ATS algorithm are studied.Using the precision of virtual clock instead of the precision of common node clock and through the ATS protocols to improved the synchronization accuracy of the network;Using hierarchical clock synchronization algorithm based on ATS protocols to improve the convergence speed of the network.The effectiveness of the improved ATS protocols is verified by MATLAB simulation.