系统工程与电子技术

• 通信与网络 • 上一篇    下一篇

无线传感网多模簇维护机制与算法

胡向东1, 徐慧芬2, 王凯1   

  1. (1. 重庆邮电大学自动化学院, 重庆 400065;
    2. 重庆邮电大学通信与信息工程学院, 重庆 400065)
  • 出版日期:2015-09-25 发布日期:2010-01-03

Mechanism and algorithm of multimode clusters maintenance for#br# wireless sensor networks

HU Xiangdong1, XU Huifen2, WANG Kai1   

  1. (1. College of Automation, Chongqing University of Posts and Telecommunications, Chongqing 400065, China;
    2. School of Telecommunication and Information Engineering, Chongqing University of
    Posts and Telecommunications, Chongqing 400065, China)
  • Online:2015-09-25 Published:2010-01-03

摘要:

无线传感网因节点众多、应用场景多样和资源约束严格而具有复杂的分布特征、组网灵活性和高度动态性,基于簇的网络管理模式面临着如何实现针对不同的簇受损情形的高效簇维护问题。建立了一种可以适应不同簇维护需要的多模簇维护机制,以自适应局域按需簇维护为目标,根据网络中簇的动态变化时间和空间进行分类簇维护,将簇维护转化为局部性和事件触发的操作,触发机制包括簇头剩余能量低于设定阈值、簇头负载不均、新节点加入或簇内节点失效退出等,基于单个簇或相邻簇的状态变化关系启动单簇(簇内)维护或多簇(簇间)维护方案。基于NS2的仿真结果表明:与传统周期性、全网重新成簇的维护模式相比,该机制能够明显降低网络维护开销,有效均衡网络节点能耗、增加传输的数据包总量、100%节点存活时网络的平均生命周期最高可延长1.8倍。

Abstract:

Due to numerous nodes, diverse scenarios and limited resource, wireless sensor networks are featured in complicated distribution, flexible networking and highly dynamic state, the clusterbased network management is facing how to energyefficiently maintain the structure of clusters for different situations. A multimode cluster maintenance mechanism is proposed to meet diverse demands of maintenance of clusters, which focuses on the target of adaptive, local and ondemand maintenance of clusters, the operations of maintenance of clusters are done according to the dynamically changed network in time and space fields, it transforms the maintenance of clusters into local and eventtriggered operations, the triggers include the residual energy of cluster head lower than a set threshold, the outofbalance loads of cluster heads, new node joining a cluster or disable nodes, etc. The program of maintenance for single cluster (innercluster) or multi clusters (interclusters) 〖JP2〗is initiated according to the damaged single cluster or neighboring ones. The results of simulation based on NS2 show that the proposed methods can significantly reduce cost of network maintenance, effectively balance the energy consumption among nodes, increase the sum of transmitted data and prolong lifetime of the network 1.8 times at most under 100% survived nodes compared with the traditional periodic wholenetwork reclustering method.