Based On The Clustering Of Wireless Sensor Network Time Synchronization Algorithm

The advancement of sensors, modern network and wireless communications technologies has promoted the emergence and development of modern wireless sensor networks. Wireless sensor networks are constructed by a large number of self-organization sensor nodes, which interact with each other and connect the physical information of the objective world with transmission network together by way of single hop or multi-hop through collecting the physical parameters within the monitoring region. Wireless sensor networks extend the ability to obtain information and have great broad application prospects, which have been attracted by academia and industry at home and abroad. The time synchronization of wireless sensor network involves two aspects which are physical time and logical time. Data obtained by sensors must contain exact time and location information, otherwise, the collected information is incomplete. The study of time synchronization algorithm is one of the key approaches in sensor networks field.Focusing on two issues of wireless sensor networks, which are energy constraint problem and time synchronization accuracy problem, this paper described the basics of wireless sensor networks, the types of time synchronization algorithm and the time synchronization mechanism of typical algorithm and analyzed the design principles an the advantages and disadvantages of each algorithm. Comparing the synchronization principles of different time synchronization algorithms, this paper summarized the advantages and disadvantages, the synchronization accuracy and the synchronization consumption. Directed at the uneven problem of energy consumption for network nodes, a time synchronization solution based on clustering has been proposed and the correctness and effectiveness of which has been verified by theoretical analysis and simulation.The main work and innovations are as follows:1.Comparing the typical time synchronization algorithm of wireless sensor networks, this paper gave the mathematical model of for RBS algorithm, TPSN algorithm and DMTS algorithm respectively. Because traditional Internet time synchronization approach doesn't apply to wireless sensor networks, this paper summarized the status of current researches on time synchronization of wireless sensor networks and investigated typical RBS, TPSN and DMTS algorithm based on receiver-receiver synchronization, pair synchronization and sender-receiver synchronization mechanisms to express the synchronization accuracy and consumption as mathematical formulas and proved the performance of different algorithms theoretically.2.Directing at the restriction of energy consumption for network nodes, CTSA algorithm and MCTSA algorithm have been proposed which are energy efficient and based on clustering. With high energy requirements of wireless sensor networks and to achieve the highest synchronization accurancy, it can guarantee the generation level depth as small as possible. Focusing on the needs of synchronization accuracy and synchronization energy consumption when designing TSA algorithm and MCTSA algorithm. On the basis of the clustering, an independent system is established inner-cluster. In the system, the nodes are in order of one-way synchronization and cluster heads can monitor the progress of synchronization in the real-time to achieve energy savings and robust pattern of synchronous collaboration. The proposed algorithms have been certified to reduce communication costs effectively and to save node energy under the premise of accuracy guaranty by the constructed mathematical model.3.The two algorithms have been simulated by MATLAB and NS2. The results of simulation demonstrate obtain the optimal data of the parameters in the network. Besides, we have simulated and verified synchronization accuracy, synchronization spending and synchronization energy consumption. The experimental data showed that the two proposed algorithms met the application requirements and performance indicators of wireless sensor networks. Therefore, the correctness and superiority of the algorithms have been confirmed.