无线传感器网络能耗优化机制研究

发布时间：2018-07-13 07:32

【摘要】：无线传感器网络(Wireless Sensor Networks,WSN)是一门面向应用的,新兴的多交叉学科,未来必然对人类的生活产生深远的影响。在WSN中,能耗问题一直是研究的重点。大量的能耗一般是在节点的通讯过程中产生的,能耗的大小和节点之间的距离以及传输的数据量有关。针对这两个方向,目前已经有了大量相关的研究工作。其中比较重要的是以LEACH协议为代表的WSN分簇路由协议以及基于代表式的数据融合算法。这些协议及算法为研究WSN能耗问题提供了重要的解决思路,但是这两种协议在减少网络能耗问题方面都存在着缺陷,本文主要对这些缺陷和改进方向做了描述,本文的主要工作包括:1.首先,针对LEACH协议存在的缺陷:随机选举簇头、节点采用单跳传输以及孤立节点信息不上传,提出了LEACH协议和BIRCH算法相结合的LEACH-B路由协议,利用BIRCH算法中的成簇半径大小和簇内子节点数来约束WSN成簇的大小。同时,每次选取最大剩余能量的节点为簇头,保证网络的稳定性;孤立节点单独成簇,采集的信息需要单独上传,保证了网络采集信息的完整性;节点之间采用多跳协同通信,减少了传输的能耗。最终达到了降低网络传输能耗,有效延长网络生命周期的目的。2.其次,针对以往的代表式数据融合方法在选取代表节点方面存在的节点代表性不强的问题,本文提出了PCCDNCD数据融合算法。在PCCDNCD算法中,节点的相关性以节点采集数据的皮尔森相关系数、节点之间的物理位置以及邻居节点的数量来衡量,提高了节点相关性的精度。然后,按照相关度对节点进行分类:sink节点、普通节点和孤立节点。不同节点选择不同的传输方式,传输或者不传输数据。实验和仿真结果表明:PCCDNCD算法融合过后的数据较DDCD和PCC算法的融合精度高,网络的剩余能量也较DDCD和PCC有了较大改善。PCCDNCD算法可以有效的延长网络的生命周期。3.设计并实现了无线传感器网络能耗优化系统EOS,文中以森林防火预警为背景,考虑将LEACH-B和PCCDNCD算法进行结合,运用到森林防火预警系统中。将程序代码烧写到100个节点中,节点部署在100*100m的区域进行实验。基站监测节点收集的温度数据及节点的剩余能量状态,通过对比一般情况下的节点的剩余能量和本文提出的EOS的剩余能量,实验结果表明,本文提出的EOS系统可以有效的促进网络平衡,减少网络能耗。
[Abstract]:Wireless Sensor Networks (WSN) is an application-oriented, emerging interdisciplinary subject, which will have a profound impact on human life in the future. In WSN, energy consumption is always the focus of research. A large amount of energy consumption is usually generated in the communication process of nodes. The size of energy consumption is related to the distance between nodes and the amount of data transmitted. In view of these two directions, there has been a lot of related research work. Among them, the WSN clustering routing protocol represented by Leach protocol and the data fusion algorithm based on representative are more important. These protocols and algorithms provide an important way to solve the problem of WSN energy consumption. However, both protocols have defects in reducing network energy consumption. This paper mainly describes these defects and the direction of improvement. The main work of this paper includes: 1. First of all, aiming at the defects of Leach protocol: random election of cluster head, single hop transmission of nodes and non-upload of isolated node information, a LEACH-B routing protocol combining Leach protocol and Birch algorithm is proposed. The size of WSN clusters is constrained by the radius of clusters and the number of sub-nodes in the Birch algorithm. At the same time, the nodes with the largest residual energy are selected as cluster heads each time to ensure the stability of the network; isolated nodes are clustered separately and the collected information needs to be uploaded separately to ensure the integrity of the information collected in the network; the multi-hop cooperative communication is adopted among the nodes. Reduced transmission energy consumption. Finally, the goal of reducing network transmission energy consumption and effectively prolonging network life cycle is achieved. Secondly, aiming at the problem that the former representative data fusion methods are not representative in selecting representative nodes, a PCCDNCD data fusion algorithm is proposed in this paper. In the PCCDNCD algorithm, the correlation of nodes is measured by the Pearson correlation coefficient of data collected by nodes, the physical position of nodes and the number of neighbor nodes, which improves the accuracy of node correlation. Then, according to the correlation degree, the nodes are classified as: sink node, common node and isolated node. Different nodes choose different modes of transmission, transmit or do not transmit data. The experimental and simulation results show that the fusion accuracy of the fusion data of the PCCDNCD algorithm is higher than that of the DDCD and PCC algorithms, and the residual energy of the network can be improved greatly compared with the DDCD and PCC algorithms. PCCDNCD algorithm can effectively extend the lifetime of the network. A wireless sensor network energy consumption optimization system (EOSS) is designed and implemented in this paper. Based on the background of forest fire prevention early warning, the combination of LEACH-B and PCCDNCD algorithm is considered to be applied to forest fire prevention early warning system. The program code is burned to 100 nodes, and the nodes are deployed in the 100m area. The temperature data collected by the node and the residual energy status of the node are monitored by the base station. By comparing the residual energy of the node under normal circumstances with the residual energy of the EOS proposed in this paper, the experimental results show that, The proposed EOS system can effectively promote network balance and reduce network energy consumption.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP212.9;TN929.5

【参考文献】