国家自然科学基金(61170219); 重庆市自然科学基金(CSTC2013jcyjA40002); 重庆市高等学校优秀人才支持计划(65)
基于无线传感网的物联网感知层传统的“全网”、“周期性”重新成簇的簇维护模式因超范围过度维护,存在维护成本高、能量浪费严重、服务全面中断、响应不及时等缺点.局域按需簇维护方法(local and on-demand maintenance of clusters,简称LDMC)将簇维护操作控制在簇受损的时间和空间范围内,通过设置触发源、预处理和维护动作分别解决簇维护启动、簇维护方式和簇维护范围问题,不仅能够克服簇更新周期确定的困难,而且可在节点失效和新节点加入时对网络拓扑和路由变化及时进行响应,减小突发事件对网络功能的影响,改善网络的稳定性并降低其维护开销.基于NS2仿真平台,分别从能量消耗、数据传输、负载平衡和突发事件响应等角度对该方法进行了测试对比,仿真结果表明,该方法能够明显减少簇维护的能量消耗、延长网络生存时间,并增加传输数据包的 总量.
The sensing layer of the Internet of things based on wireless sensor network requires intensive maintenance over the involved nodes and beyond by the conventional whole-network and periodic model of cluster maintenance. It therefore results in some shortcomings such as high cost of maintenance, heavy waste of energy, full service interruption, and delayed response to incident. This paper proposes a new method, namely, local and on-demand maintenance of clusters (LDMC), to carry out the operations of maintenance only in a restricted range of time and space which is decided by the damaged clusters. LDMC resolves issues such as when to start, which type or who to be maintained by the triggers, pre-processing or different operations of maintenance. The presented scheme is helpful not only to make critical decision on the timing of cluster update, but also to provide in-time response to the change of topology or route of network with the disabled or new nodes. As a result, it allows to reduce the impact of incidents on the function of network, to improve the stability of network, and to cut down the cost of maintenance. Comparison tests are performed on consumption of energy, transmission of data, balance of load and response to incident based on NS2 simulation platform, and the results suggest that the proposed method is able to significantly reduce energy consumption in maintenance of clusters, to prolong the lifetime of network, and to increase the total amount of the transferred data packets.