摘要
针对随机分簇算法未考虑节点位置和对节点能量利用不充分的问题,提出了一种基于改进萤火虫聚类的异构无线传感器网络能耗优化路由算法(IFCEER)。该算法将改进萤火虫聚类算法用到高能节点分簇中,在时间充裕的数据传输阶段预测与聚类中心和基站等位置相关的主副簇头,进而形成结构紧密的全局最优簇集合,避免簇头可能集中于局部区域造成簇半径随意扩大的缺点,平衡了异构节点的能耗,降低了频繁重新聚类消耗能量的风险。仿真实验结果显示:与原有算法相比,在自由空间模型主导的100 m×100 m监测环境和多路径衰减模型主导的300 m×300 m监测环境中,网络内第1个节点死亡时间分别延迟43%到225%;随着节点间传输距离的增加,300 m×300 m监测环境能耗减少达到60%。
To solve the problem that the position of nodes does not be considered and the energy utilization of nodes is not sufficient in stochastic clustering algorithms,an optimized energy efficient routing algorithm based on improved firefly clustering is proposed for heterogeneous wireless sensor networks.It uses the improved firefly clustering algorithm to cluster high energy nodes,predicts the main and vice cluster-heads related to the position of clustering center and base station in a time abundant data transmission phase,and then forms a tightly structured global optimal cluster set.These strategies not only avoid the disadvantage that cluster heads may be concentrated in a local area which causes the cluster radius to expand arbitrarily,but also balance the energy consumption among heterogeneous nodes and reduce the risk of frequent re-clustering.The simulation results showed that compared with former algorithm,in the 100 m×100 m monitoring environment dominated by the free space channel model and the 300 m×300 m monitoring environment dominated by the multipath fading channel model,the death time of the first node in the network was delayed by 43%to 225%,respectively.With the increase of transmission distance between nodes,the energy consumption in the monitoring environment of 300 m×300 m was reduced by 60%.
作者
罗剑
毕晓东
LUO Jian;BI Xiao Dong(School of Digital Information,Zhejiang Technical Institute of Economics,Hangzhou 310018,China)
出处
《传感技术学报》
CAS
CSCD
北大核心
2018年第10期1584-1591,共8页
Chinese Journal of Sensors and Actuators
基金
国家自然科学基金项目(61472363)
浙江省科技厅项目(2015C31107)
浙江省科技厅项目(2016C33109)
关键词
异构无线传感器网络
改进萤火虫聚类
高能节点
预测分簇
能耗优化
heterogeneous wireless sensor networks
improved firefly clustering
high energy node
predictive clustering
optimized energy consumption