WSN中基于多移动代理的路径规划及数据融合研究

发布时间：2018-01-15 02:13

本文关键词：WSN中基于多移动代理的路径规划及数据融合研究　出处：《南昌航空大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着通信技术,传感技术等技术的日益发展,科研学者研制出了各种传感设备,由众多的无线传感设备,以无线通讯的方式组合而成的网络称为无线传感网络(WSN)。因为其体积小,无线通信等优越前提使其被普遍应用于农业、医疗、环境及交通等范畴。WSN在传感节点分布密集,范围广阔的情况下由于频繁的通信将带来网络负载过大,监测区域重叠导致传感数据冗余等现象。如何在WSN中提高数据融合性能如:降低节点能耗,缩短信息传输延时,提高数据融合精度等是WSN急需克服的问题。由于移动Agent有着很好的自主性和独立判断能力,将其应用于WSN的数据融合之中可以提高诸多数据融合相关性能。但移动Agent的性能与其在网络中的访问路径息息相关,如何为移动Agent设计合适的访问路径并结合数据融合技术达到最优的性能是急需解决的问题。本文在移动Agent的基础之上提出在WSN中基于多Agent的路径规划及数据融合策略,其主要研究内容如下:(1)提出多Agent基于异常网络环境的路径规划策略。根据网络规模,分配不同数量的移动Agent。每个移动Agent根据网络环境进行初始化操作,并依据节点间的开销和节点奇异比规划一条高效的节点访问路径。(2)提出轻量级数据融合机制。数据融合策略分为:单个移动Agent数据融合和全网多个Agent数据融合。同时定义奇异值评判标准并据此找出与删除,将余下数据实行融合操作得到可信可靠的结果。(3)改造网络仿真环境。由于本文在ONE实验平台下进行实验,并且其本身并不支持移动Agent方式进行实验,因此从节点,通信机制等方面对该实验平台进行改造,使其满足移动Agent的数据融合实验。(4)实验结果对比分析。从网络传输延时,能量消耗和网络节点死亡率三大性能指标在不同网络环境下对LCF、GCF和本文算法BFSR进行比较分析。实验表明,本算法在异常网络环境下有着更优的优越性。(5)算法复杂度对比分析。针对LCF、GCF及本文BFSR算法分别根据算法时间和空间复杂度两个角度计算与比较。
[Abstract]:With the development of communication technology, sensing technology, researchers developed a variety of sensing devices, by many wireless sensor equipment, by the way of wireless communication and network is called wireless sensor network (WSN). Because of its small size, the advantages such as the premise of wireless communication is widely used in agriculture, medical, environmental and traffic areas such as.WSN in the sensor nodes are densely distributed, a wide range of circumstances because of the frequent communication will bring the network load is too large, resulting in the monitoring area overlap sensor data redundancy. How to improve the performance of data fusion in WSN such as: reduce energy consumption, reduce the information transmission delay, improve the accuracy of data fusion WSN is in urgent need of overcoming problems. Because the mobile Agent has good autonomy and independent judgment, which is applied in the WSN data fusion can improve many data fusion Close the performance. But the mobile Agent performance and in the network access path is closely related to the access path of mobile Agent and combined with the design of appropriate data fusion technology to achieve optimal performance is an urgent problem. This article is based on the mobile Agent proposed path planning and data fusion strategy in WSN based on multi Agent. The main research contents are as follows: (1) proposed the multi Agent path planning strategy based on network environment. According to the abnormal distribution of Agent. mobile network scale, each mobile Agent with different number of initialization operation according to the network environment, and on the basis of inter node overhead and node singular than planning an efficient access path node (2). Proposes a lightweight data fusion mechanism. Data fusion strategy is divided into: single mobile Agent network data fusion and multiple Agent data fusion. At the same time the definition of singular value criterion and according to Select and delete the remaining data, implement credible results fusion operation. (3) the transformation network simulation environment. The experiment was carried out under the ONE platform, and it does not support mobile Agent by way of experiment, so from the node communication mechanism for the transformation of the experimental platform, which can meet the mobile Agent data fusion experiments. (4) comparative analysis of experimental results. The network transmission delay, energy consumption and network node mortality of three performance indicators in different network environment of LCF, GCF and BFSR algorithm are compared and analyzed. Experimental results show that this algorithm has advantages of better in abnormal network environment. (5) analysis of the complexity of the algorithm. Compared to LCF, compared with GCF and the BFSR algorithm respectively according to the time and space complexity of the algorithm two angle calculation.

【学位授予单位】：南昌航空大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP212.9;TN929.5

【相似文献】