多信道无线传感器网络组网与定位方法研究
文献类型：学位论文
| 作者 | 肖金超
|
| 学位类别 | 博士 |
| 答辩日期 | 2014-05-30 |
| 授予单位 | 中国科学院沈阳自动化研究所 |
| 导师 | 于海斌 ; 曾鹏 |
| 关键词 | 无线传感器网络 多信道 组网 定位 调试 |
| 其他题名 | Network Initialization and Localization Approaches to Multi-Channel Wireless Sensor Networks |
| 学位专业 | 机械电子工程 |
| 中文摘要 | 无线传感器网络是近年来兴起的一项新型信息获取技术，在工业、军事、环境、医疗、家庭等众多领域有较高的应用价值和广阔的应用前景。前期的研究主要考虑传感器节点的能量限制，关注低开销网络协议设计，为了降低组网复杂性，传感器节点一般在单一、固定的信道上通信。 近年来，随着工业应用的兴起，人们开始关注无线传感器网络的传输可靠性、实时性等性能指标。由于单信道通信在抗干扰、信道容量等方面存在缺陷，已经不能满足工业应用的要求。而可以在多个信道进行灵活通信的无线传感器网络技术逐渐成为研究的热点，现有的IEC工业无线传感器网络标准WirelessHART和WIA-PA都采用了多信道无线传感器网络技术。 多信道无线传感器网络虽然具有高容量、抗干扰、高可靠的优势，但由于通信信道的不确定，使得组网的复杂度极大增加，如何在能量受限的条件下完成组网成为研究的主要难点之一；随着通信信道数量的增加，网络可接入节点的数量增大，如何实现大规模节点快速、高效的组网是另一个难点；由于多信道通信具有频率选择性衰减等特征，如何利用多信道通信的差异性实现高精度定位开始成为研究的热点；此外，针对通信信道灵活变化、通信过程中交互复杂的特点，如何提高网络调试效率成为多信道无线传感器网络开发必须解决的难题。 本文针对多信道无线传感器网络的挑战问题，以多信道无线传感器网络的组网、定位和开发调试为主要研究内容，取得了以下创新成果： 1、针对能量严格受限条件下的多信道无线传感器网络组网问题，提出了基于报文快速捕获的低功耗组网方法。通过设计最小化报文捕获时间的线性规划，解决了通信信道不确定条件下的报文快速捕获问题；在此基础上，设计了快速时间同步和邻居发现方法，降低了组网能耗； 2、针对大规模传感器节点在组网过程中竞争信道引发的冲突问题，提出了基于扩展共享时隙的并行组网方法。通过在WIA-PA网络的超帧结构中扩展共享时隙，划分了冲突时域；在此基础上，以节点剩余能量、信道质量和启动时间为指标，设计了差异化的节点入网时间，降低了冲突概率；并通过资源预先分组方法，避免并行组网过程中的资源分配不一致问题，实现了大规模传感器节点的并行快速组网； 3、针对利用多信道通信差异进行节点定位的问题，提出了基于频率选择性衰减特征的高精度定位方法。通过现场实验数据，分析了无线电信号强度衰减与位置及频率的关系；在此基础上，估算出多信道无线电波在直线传播情况下衰减后的信号强度值，并用该值进行测距和定位，极大提高了定位精度。 4、针对多信道无线传感器网络交互式错误难以调试的问题，提出了基于报文序列分析的错误调试方法。使用基于关联规则挖掘方法对报文序列进行分析，进行反映异常交互行为的报文序列频繁项挖掘，在此基础上设计了多信道无线传感器网络调试工具。 |
| 索取号 | TP212/X47/2014 |
| 英文摘要 | Wireless sensor network is regarded as a novel information acquisition technology, which plays an essential role in a number of fields including industry, military, and environmental process in recent years. However, previous studies mainly considered the energy limitation of sensor nodes and low-expended design for network protocol, leading that sensor node could only communicate with each other in a fixed single channel to reduce network complexity. With the development of industrial applications, an increasing number of researchers incline to focus on the performance metrics of wireless sensor networks, such as transmission reliability, and real-time property. Nevertheless, single-channel communication has drawbacks in anti-interference and channel capacity, which could not meet the requirements of current industrial applications. On the contrary, multi-channel wireless sensor network is gradually drawing more attention due to its flexible communication in several channels among sensor nodes. Two existing IEC industrial standards have adopted multi-channel wireless sensor network technology. One is WirelessHART; and the other is WIA-PA. Despite of the anti-interference and high reliability of multi-channel wireless sensor network, the complexity for network initialization is also increased to a great extent because of the uncertain communication channels. Therefore, a number of difficulties need to be overcome during the network initialization process, including: (1) Initialize the networking in energy-constrained condition; (2) Process fast and efficient network initialization for large number of nodes accompanied by the increased communication channels and numerous nodes accessing to the network; (3) Utilize the communication discrimination rule among multi-channels to achieve highly accurate localization according to the frequency selective attenuation characteristic; (4) Enhance the network debugging efficiency to overcome the complex interaction features in communication process. This paper focused on the network initialization, localization, and developing and debugging aspects of multi-channel wireless sensor network. The new innovations are listed as follows. 1. A low-power network initialization approach was investigated in this paper, aiming to apply multi-channel wireless sensor network into energy limited conditions, which exhibited packet capturing based and low-energy consuming properties. The method could not only solve the fast packet capturing problems via the linear programming on the basis of minimizing the capturing time, but also decrease the energy consumption by fast time synchronization and neighbor discovery. 2. For the confliction caused by the channel competition among large-scale sensor nodes, the paper proposed a parallel network initialization method. It did not only successfully divide the conflicted domain through extending the sharing time slots in the superframe of WIA-PA network, but also selected the different network accessing time to diminish confliction probability according to residual energy, channel quality and starting time, as well as achieved the fast parallel network for a large-scale sensor nodes by the advanced resource grouping method to avoid inconsistence for resource allocation. 3. The paper also proposed a highly accurate localization method based on the frequency selective attenuation feature to achieve node localization using multi-channel communication discrimination. The experimental data were firstly analyzed to investigate the relationship between the strength of the signal attenuation and the location and frequency. On this basis, the multi-channel signal strength value could be calculated for the later ranging and localization process after the attenuation in the linear propagation process, which greatly improved the localization accuracy. 4. Finally, in order to debug the interactive faults, the paper suggested a packet sequence based debugging method which elaborately analyzed the packet sequence using association rule mining method to reflect the abnormal sequence of the interactive packets by frequent pattern mining, as well as designed debugging tools for multi-channel wireless sensor networks. |
| 语种 | 中文 |
| 产权排序 | 1 |
| 页码 | 107页 |
| 分类号 | TP212 |
| 源URL | [http://ir.sia.ac.cn/handle/173321/14822]  |
| 专题 | 沈阳自动化研究所_工业控制网络与系统研究室 |
| 推荐引用方式 GB/T 7714 | 肖金超. 多信道无线传感器网络组网与定位方法研究[D]. 中国科学院沈阳自动化研究所. 2014. |
入库方式： OAI收割
来源：沈阳自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。