埋地水管组成的自来水管网是城市重要的基础设施之一，然而由于埋地水管的不可见性，其泄漏问题一直制约着城市的健康发展和人民生活的改善。因此，对埋地水管泄漏技术的研究有着重要的意义。在埋地水管泄漏点的检测和定位技术研究中，时间同步技术是其一项支撑技术。只有建立在同步的基础上，将无线传感器网络同步协议应用于城市自来水管网中，才能及时检测出泄漏点，控制和减少埋地水管的泄漏率。 本文针对无线传感器网络埋地水管泄漏检测系统中节点间时间同步算法进行研究。阐述了埋地水管网检测与定位系统基本组成结构和网络拓扑结构等，得到该系统是网络拓扑简单的单跳网络。在此基础上，采用最小二乘法对TPSN时间同步算法进行改进，并在对IEEE802.15.4的MAC层技术进行研究，提出将执行CCA（空闲信道检测）之前的随机退避周期延时改为定值退避周期延时的优化方案。在Linux操作系统、NS2环境下采用C++和OTCL交互式编程，对该算法进行了仿真，仿真结果表明改进后算法应用在埋地水管检测系统中定位精度达到±0.208m，符合系统设计要求。该算法减小了时间同步误差，适用于网络拓扑结构固定的单跳网络。

Buried pipes is one of the city's major infrastructure, However, because of buried pipes is invisibile. buried pipes leak problem has been restricting the development of the city and improving the health of people's lives. So, Research on buried pipes leak technology is very important. In the study of detection and location of buried pipe leaks, The core technology is the detection and localization time synchronization technology. Only based on synchronization, Apply wireless sensor network synchronization protocol in city water pipe network in order to detect leaks in time, control and reduce the rate of buried pipes leak. In this paper, Study the wireless sensor network underground water pipe leak detection system between the time synchronization algorithm s. Elaborated buried pipe network detection and location system structure and the basic components such as network topology, network topology to get the system is a simple single-hop network. Again this basis using the least squares method TPSN time synchronization algorithm is improved, and on the IEEE802.15.4 MAC layer technology research, proposed the implementation of CCA (idle channel detection) cycle delay before the random backoff value to backoff cycle delay optimization. In the Linux operating system, using C + + under NS2 environment and OTCL interactive programming, the simulation results show that the improved algorithm for the detection of buried pipe systems positioning accuracy of ± 0.208m, in line with the system design requirements. To note is that the algorithm is mainly applied to the fixed network topology single-hop network.