Research On Time Synchronization In Wireless Sensor Networks Under Imperfect Conditions

Time synchronization is one of the hot issues in the research of wireless sensor networks.Compared with the centralized time synchronization protocol,distributed time synchronization protocol has better robustness and scalability.However,because there are many non ideal factors in the wireless sensor network,the application scope of the distributed time synchronization protocol is limited.This thesis mainly study the design and improvement of time synchronization protocols under the two non ideal conditions: limited memory and communication delay which cannot be ignored.First,in view of the limited memory of nodes,this thesis point out the serious problem of memory usage in Average TimeSynch(ATS).Because the distributed time synchronization algorithm requires each node stores the data information of its neighbor nodes,this kind of algorithm has great computing cost and memory cost.A memory-lite time synchronization protocol(MLTS)is proposed.The memory size used to store data in the node can be given artificially.By rational memory management and data exchange,memory constrained nodes can still be used in dense and mobile networks.This thesis analyzes the performance of the consensus time synchronization protocol ATS and MTS under the communication delay,and points out the bad effect of the jitter of drift compensation on the stability of the clock.An improved clock model is proposed to suppress the drift compensation jitter.An enhanced ATS is proposed,which optimizes the performance of ATS under communication delay.A time synchronization protocol based on the maximum and average consensus is proposed.In the case of communication delay,the stability of the drift compensation can be guaranteed.When the delay is too large,the protocol automatically degenerates into a simple offset-compensate-only algorithm,which greatly improves the reliability of the clock.In this thesis,a ZigBee network is built,and a timestamping method based on SFD capture is adopted.It eliminates most of the uncertain communication delays and greatly improves the stability of the time synchronization algorithm.Besides,the proposed time synchronization protocol is verified on the ZigBee platform.Result shows that the proposed time synchronization algorithm has good performance for memory management,drift compensation,and has high accuracy and robustness.