Study Of Wireless Sensor Network Time Synchronization Algorithm

Tremendous advances in communication, embedded computation and sensor technology have occurred in the development of wireless sensor networks, which is made by sensors, microprocessor and wireless communication interface, and have integrated collecting, processing and communication all-in-one. A huge surge of interest in this filed has been triggered by large potential in a wide range of applications such as: military surveillance; environmental monitoring; medical sanitation; physiological monitoring; smart spaces; tragedy salvation. And it also has been highly valued by academic and industrial fields.Time synchronization is a fundamental service for both design and application of wireless sensor networks. However, the complexity, energy, cost and size conflicts between traditional algorithm such as NTP and GPS and WSN prohibit their use in WSN. Because of the cost, power and size constraints of the individual nodes, there are many new challenges. Developing time synchronization methods suitable for wireless sensor networks have attracted wide attention.The research of this thesis is time synchronization in wireless sensor networks. First, the necessity, fundamental principles, existed challenges, evaluation criterion of the performance and some important in design of time synchronization are described. Then, study existing technologies of time synchronization and classify them. Some representative synchronization algorithms are discussed deeply. Based on these, a novel simple and low cost time synchronization algorithm, which is suitable for dense network, is presented. It combined unidirectional reference broadcast and pair-wise mechanism, so that only one node in broadcast domain needs time synchronization to it. Times-tamping at MAC layer make it achieves a good precision. Compared to TPSN, our algorithm makes good use of broadcast information, and reduces communication messages greatly. Simulation results show that it has a good precision and low message cost. After that, we implemented new algorithm on our test-beds. In addition to these, we also analyze time synchronization in terms of security. Finally, we summarized this paper.