高级检索
    苏金钊 刘丽艳 吴 威. 适用于周期休眠MAC协议的分簇时间同步算法[J]. 计算机研究与发展, 2010, 47(11): 1893-1902.
    引用本文: 苏金钊 刘丽艳 吴 威. 适用于周期休眠MAC协议的分簇时间同步算法[J]. 计算机研究与发展, 2010, 47(11): 1893-1902.
    Su Jinzhao, Liu Liyan, and Wu Wei. Clustering Time Synchronization Algorithm for Periodic Sleep MAC Protocol[J]. Journal of Computer Research and Development, 2010, 47(11): 1893-1902.
    Citation: Su Jinzhao, Liu Liyan, and Wu Wei. Clustering Time Synchronization Algorithm for Periodic Sleep MAC Protocol[J]. Journal of Computer Research and Development, 2010, 47(11): 1893-1902.

    适用于周期休眠MAC协议的分簇时间同步算法

    Clustering Time Synchronization Algorithm for Periodic Sleep MAC Protocol

    • 摘要: 无线传感器网络中节点能量有限,常采用周期休眠的方式工作,而周期性休眠机制的实现依赖于节点间的时间同步方法.基于竞争的周期性休眠MAC协议的典型代表是S-MAC,在S-MAC协议的时间同步算法基础上,通过引入簇控制和边界节点控制方法提出一种分簇时间同步算法,该算法适用于周期性休眠的MAC协议.仿真和物理实验表明,分簇时间同步相比S-MAC时间同步方法能够有效控制网络中的簇数和边界节点数,减少时间同步开销和端到端传输时延,从而节省能耗,延长网络生存周期.

       

      Abstract: The MAC protocol of wireless sensor network takes responsibility of allocating wireless channels. Due to the limited energy of wireless sensor nodes, they often work in the way of periodic sleep. However, periodic sleep leads to increase in transmission delay. The realization of periodic sleep depends on time synchronization methods between nodes. A typical representative of competition-based periodic sleep MAC protocol for wireless sensor network is S-MAC. On the basis of the time synchronization algorithm of S-MAC, we propose a clustering time synchronization algorithm by introducing cluster and border node control methods, redesign the format of MAC frame, present the procedure of building cluster, and study the producing, controlling and replacing strategies of border node. This algorithm is suitable for periodic sleep MAC protocols. Results of simulation and practical experiments show that compared with S-MAC, this algorithm can significantly control the numbers of cluster and border node, decrease the overhead of time synchronization and end-to-end transmission delay, so as to save nodes energy and extend lifecycle of the whole network. In a competitive unicast situation, the network throughput is up to 208.66Bps which can satisfy the data transfer requirement of general sensor network. The cost of time synchronization control can be reduced by 35%, and multi-hop transmission delay decreased by 4%.

       

    /

    返回文章
    返回