一种基于数据驱动的无线传感器网络分簇算法  

A CLUSTERING ALGORITHM BASED ON DATA-DRIVEN FOR WIRELESS SENSOR NETWORKS

作　　者：衣晓[1] 邓露[1] 刘瑜[1] 

机构地区：[1]海军航空工程学院信息融合技术研究所,山东烟台264001

出　　处：《计算机应用与软件》2011年第8期71-75,共5页Computer Applications and Software

基　　金：国家自然科学基金(60572161);全国优秀博士论文作者专项资金(200443);"泰山学者"建设工程专项资金

摘　　要：由于无线传感器网络的能量有限,尽可能地减少能量消耗成为设计者的主要研究目的之一。在研究LEACH及其改进算法TEEN等算法的基础上,提出一种基于数据驱动的无线传感器网络分簇算法,即CADD算法。在CADD算法中,节点的数据发送基于两个参数:采集数据的信息量以及节点没有发送数据时延,通过两个参数的计算结果与设置阈值的比较关系判断数据的发送与否。在数据传输过程中,提出使用数据传输状态与节点信息素转换的方法,用于减少数据传输冲突。此外,在簇头选举阶段,通过使数据传输量小的节点较大概率地成为簇头,有效优化了簇头选举。最后,通过仿真的方法,将CADD算法与LEACH算法和TEEN算法进行对比,结果表明CADD算法优于LEACH算法和TEEN算法。Due to the limitation of energy in wireless sensor networks,reducing energy usage as much as possible becomes one of the main researching aims for designers.Based on the algorithm LEACH and its improved algorithm TEEN,a clustering algorithm based on data-driven for wireless sensor networks（CADD） is presented,in which the nodes judge whether or not to send data by two factors： information of the data collected and without data-delay when nodes transmitting,whether the data would be transferred or not is judged through the comparison of relation between the calculation results of two factors and the set threshold value.During the data transmission,a converting method between data transfer state and node information is presented in the paper to reduce data transmission conflict.Additionally,making those nodes sending few data have more chances to be the cluster heads in cluster heads election stage could effectively optimise the cluster heads election.Finally,the authors compare CADD with LEACH and TEEN by simulation,and find from the results that CADD shows better performance than that of LEACH and TEEN.

关 键 词：无线传感器网络 数据驱动 信息量 数据时延 簇头 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]