无线网络数据采集与数据汇聚算法研究

发布时间：2018-07-05 11:50

本文选题：无线传感器网络 + 数据采集　；参考：《曲阜师范大学》2017年硕士论文

【摘要】：近年来,无线传感器网络(WSNs)由于具有低成本、自组织、动态拓扑、多跳路由等特点,现已成为许多重要应用中的首选解决方案,例如:入侵检测、目标追踪,实现工业自动化等。无线传感器网络中的传感器节点能够感知区域周围环境中的信息数据,而当这些节点将感知到的数据传送到终端的时候,就出现了数据采集这一技术,然而,同一区域的数据可能会被若干个传感器检测到,那么传感器节点就会采集到重复的数据,另外,对于同一个地区的数据来说,如果被多次采集到也会消耗很多的传感器节点的能量,而且有时候很多应用需要的并不是所有的原始数据而是一些特定的数值,因此,又出现了数据汇聚这一技术。在无线传感器网络中,数据采集和数据汇聚是基础但又很重要的操作。本文研究的是汇聚延迟问题,即保证传感器节点发送的数据在被接收节点正确接收的情况下,尽可能将所用时间缩短,以此保证数据传输的时效性。在无线传感器网络中,主要常用两种干扰模型:一是协议干扰模型,二是物理干扰模型(SINR),本文对这两种干扰模型都进行了研究,并分别提出了不同的数据汇聚算法,得到了较好的结果。在无线传感器网络中由于许多概率性损耗链路的存在,在现实的概率性网络模型下获得一个数据汇聚树变得更为实际,因此,针对这个问题,我们提出了一个数据汇聚树的创建算法,与之前的工作相比,该算法能够保证具有更高的传输成功效率,此外,我们也研究了基于生成的数据汇聚树上的数据汇聚,并得到算法可以在有限轮之内完成数据汇聚。针对物理干扰模型下的数据汇聚问题,本文主要是采用网格划分和休眠机制来避免干扰以及减少能量消耗。首先,我们创建了一棵在SINR模型下的数据汇聚树,其次,采用网格划分的思想,提出了汇聚链路调度算法,最后,我们结合创建的数据汇聚树以及链路调度算法,加入节点休眠机制,完成数据汇聚算法DA。此外,我们还提出了一个改进的数据汇聚调度算法IDDA,算法的基本思想与DA类似,改进的不同就是网格划分的标准不同,并且IDDA算法可以以分布式方式执行,算法在执行链路调度的时候采用的是分簇的思想,最后通过理论分析证明DA算法的有效性并通过仿真实验比较得到IDDA的延迟比现有的算法DAS延迟更小。
[Abstract]:In recent years, wireless sensor networks (WSNs) have become the preferred solutions in many important applications, such as intrusion detection, target tracking, due to their low cost, self-organization, dynamic topology, multi-hop routing and so on. Realize industrial automation etc. Sensor nodes in wireless sensor networks can sense information data in the surrounding environment of the region, and when these nodes transmit the perceived data to the terminal, the technology of data acquisition appears, however, Data from the same area may be detected by several sensors, so the sensor node will collect duplicate data, and for the same area, If it is collected many times, it will consume a lot of energy of sensor nodes, and sometimes many applications need not all the original data but some specific values, therefore, the technology of data convergence appears again. In wireless sensor networks, data acquisition and data aggregation are basic but important operations. In this paper, the convergence delay problem is studied, that is, to ensure that the data sent by the sensor node is received correctly, the time used is shortened as much as possible, so as to ensure the timeliness of the data transmission. In wireless sensor networks, two kinds of interference models are commonly used: one is protocol interference model, the other is physical interference model (SINR). Good results have been obtained. Because of the existence of many probabilistic lossy links in wireless sensor networks, it becomes more practical to obtain a data aggregation tree under the real probabilistic network model. We propose an algorithm for creating data aggregation tree. Compared with previous work, this algorithm can guarantee higher transmission efficiency. In addition, we also study the data aggregation based on the generated data aggregation tree. And the algorithm can complete the data aggregation in the limited wheel. To solve the problem of data aggregation under physical interference model, this paper mainly uses meshing and hibernation mechanism to avoid interference and reduce energy consumption. First of all, we create a data aggregation tree based on SINR model. Secondly, using the idea of grid division, we propose a convergence link scheduling algorithm. Finally, we combine the data aggregation tree and the link scheduling algorithm. Add the node dormancy mechanism to complete the data aggregation algorithm DA. In addition, we propose an improved data aggregation scheduling algorithm, IDDA. the basic idea of the algorithm is similar to DA, the improvement is different from the standard of grid partitioning, and IDDA algorithm can be implemented in a distributed manner. The algorithm adopts the idea of clustering in the execution of link scheduling. Finally, the validity of DA algorithm is proved by theoretical analysis. Compared with the existing algorithms, the delay of IDDA is smaller than that of existing algorithms.
【学位授予单位】：曲阜师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP212.9;TN929.5

【参考文献】