【作者】 刘丽萍；

【导师】 孙优贤； 王智；

【作者基本信息】 浙江大学 ， 控制科学与工程， 2006， 博士

【摘要】 无线传感器网络是由大量无处不在的、具有无线通信与计算能力的微小传感器节点构成的自组织分布式网络系统，是能根据环境自主完成指定任务的智能系统；是随着无线通信和嵌入式计算技术、传感器技术、微机电技术的发展而发展起来的一种新兴信息获取技术，改变了人类与自然界的交互方式；在军事、环境监测、空间探索、反恐防暴、灾难援救、医疗卫生、智能家居等方面显示了潜在的巨大应用价值。网络覆盖是无线传感器网络研究和应用的关键性基础问题，是其他网络研究的基础，直接影响着网络的使用性能。无线传感器网络超大规模，节点自由密集分布、能量受限、通信能力受限、计算和存储能力受限的特点和无人职守、恶劣的应用环境给网络覆盖带来了很大的挑战。本文从自由分布节点覆盖效率最大化、节点协同覆盖、多目标关联覆盖和适应被监测区域物理量分布特性的网络覆盖四个方面研究了无线传感器网络的覆盖问题。论文主要的研究内容和创新点包括以下几个方面。(1)综述了无线传感器网络覆盖现有的研究成果。分析了无线传感器网络覆盖面临的问题，系统的提出了网络覆盖的评价体系，归纳了网络覆盖设计中需要注意的问题，为后面的研究做了准备。(2)针对移动节点自由分布网络覆盖中，覆盖效率低下的问题，研究了理想状态下的覆盖效率最优节点分布问题，推导出同构节点的最优分布；以理论研究结果为指导，考虑无效移动、边界效应和最佳移动距离等因素，改进了虚拟力算法VFA，提出了高能效的虚拟力算法CEVFA。(3)为满足被监测区域多个目标不同覆盖质量的需求，在CEVFA的基础上，提出了特殊区域优先覆盖的目标多重覆盖算法WMCA；该算法在保证区域完整监测的情况下能够满足对特殊关注点的覆盖重数要求，而且能量消耗量少；算法计算量小，适合资源有限的无线传感器网络；是目前为解决区域中特殊目标点多重覆盖而提出的第一个节点覆盖算法。(4)考虑到某些应用中，单个节点难以独立完成对目标对象的覆盖，我们提出了节点协同覆盖的思想；详细推导并分析了协同覆盖概率、节点数目、节点参与协同覆盖的最低覆盖概率之间的关系；分析了协同覆盖算法在不同节点分布情况下的性能；该方法可以有效利用网络资源提高感知质量；在协同覆盖模型的基础上，设计了能量自适应的协同覆盖优化算法CTCO；仿真结果表明，该算法在改善网络覆盖概率的同时，延长了网络的使用寿命。(5)通过分析能量受限节点和障碍造成的节点协同工作瓶颈问题，研究建立了节点能耗与网络覆盖质量之间的关系，节点协同覆盖质量和移动、监测能量消耗的数学优化模型，并且进行了求解和仿真分析，提出了移动节点网络中的协同目标覆盖优化算法CTCOMSN；对静止节点难以完成覆盖任务和被监测区域存在障碍的应用场景进行了算法仿真，结果显示移动节点的协同覆盖优化算法不仅可以提高覆盖效率，而且对于延长网络使用寿命也是很有效的；该算法也可以应用在一般的网络覆盖中。(6)在多目标监测的网络覆盖中，一个目标处于多个传感器节点的感知范围内，而一个传感器节点有时也会同时覆盖多个目标，这样目标与目标之间，节点与节点之间，存在着一定的关联。从这些关联入手，设计了多目标关联覆盖算法MTACA，利用数据挖掘中的关联规则挖掘方法确定目标集合和传感器节点集合，通过节点集合的工作状态转换完成目标的完全覆盖，并且达到延长网络使用寿命的目的；对算法进行了仿真，并与改进的目标覆盖PEAS算法进行了比较；结果表明MTACA算法在目标完全覆盖能力和网络使用寿命上明显优于改进PEAS算法。本章的另外一个重大贡献是首次从目标和节点的关联关系入手研究网络覆盖问题，对以后网络覆盖及其它研究是一个很大的启示。(7)在大规模的环境监测应用中，考虑被监测区域物理量场分布特性和传感器节点的位置关系，将计算流体力学中的嵌套网格技术引入无线传感器网络覆盖的算法设计中，提出适应物理量分布的嵌套网格覆盖方法NGECA；这也是在首次从被监测物理量分布的角度设计网络覆盖方法，对于无线传感器网络其他方面的研究有很大的指导意义；通过仿真研究了空间步长、物理梯度等参数对算法性能的影响；同时设计了二维应用环境中嵌套网格覆盖算法，仿真结果显示NGECA算法可以利用较少的传感器节点提供感知精度很高的被监测对象信息，在节约网络资源的同时，增加了网络的鲁棒性和稳定性。(8)考虑长期监测环境应用中的时间因素，提出适应时间变化的实时NGECA算法，满足应用的实时性要求；仿真结果表明物理量发生变化后，该算法部署的节点能够迅速提供变化区域的详细信息；该算法是一种准三维的节点部署方法，为以后无线传感器网络实时特性需求的应用提供一些参考。更多还原

【Abstract】 Wireless Sensor Networks (WSNs), which are made viable by the convergence of micro-electro-mechanical system technology, wireless communications and digital electronics, have changed the way that human recognize and communicate with the physical nature. WSN is a self-organized distributed intelligent system comprising low-cost, low-power, multifunctional sensor nodes that are small in size and capable of wireless communication tethered in short distances. These tiny sensor nodes, which consist of sensing, data processing, and communicating components, leverage the idea of sensor networks based on collaborative effort of a large number of nodes. As a novel mode of computing and a hotspot of information technology after internet, Wireless Sensor Networks promise many new application areas, such as military applications, environmental applications, health application, home application and other commercial application.Coverage is vital and also fundamental in wireless sensor networks’ research and applications, and has great influence on performance of the networks. Characteristics of sensor nodes such as large-scaled and dense deployment, constraints on power, communicating ability, computing ability and memory ability, and unattended infertile circumstance bring great challenges in network coverage. Focusing on the most efficient coverage of random distributed sensor nodes, collaborative coverage, multiple targets associative coverage and coverage adapting to the application physical nature, the dissertation studies on the coverage problems in wireless sensor networks. The major contribution of this dissertation is specifically stated as follows.(1) Research on wireless sensor networks coverage is surveyed, while the challenge in WSN coverage is analyzed. Network coverage evaluation system is proposed and design issue in WSN coverage is concluded.(2) Focusing on the coverage efficiency, the ideal optimal distribution is deduced in mobile sensor networks coverage. According to the ideal distribution, coverage-efficient coverage algorithm CEVFA is presented which is based on virtual force algorithm VFA, considering useless move, boundary and optimal moving distance etc.. Simulation results show that the performance of CEVFA is better than VFA.(3) In order to satisfy the different requirement of different interest point synchronously, a new coverage algorithm WMCA is proposed. WMCA is fit for WSN for its energy-efficiency and low computing complexity. It is the first coverage algorithm that providing different coverage service for different interest points synchronously till now.(4) In some application, a sensor node can not cover the object independently. Acollaborative coverage manner is presented to fulfill the application that only one sensor node cannot. Relationships among system collaborative coverage probability, the number of sensor nodes and the lowest coverage probability that nodes have to work are deduced in detail. The performance of collaborative coverage is analyzed with different sensor nodes distribution. An energy-adaptive collaborative coverage optimization algorithm CTCO is proposed, while the simulation results show that CTCO is able to improve the coverage quality and prolong the networks lifetime.(5) Mobile sensor nodes are introduced to obstructed and energy limited WSN coverage. Optimal relationship model is built considering collaborative coverage quality, moving distance and energy cost. Collaborative targets coverage optimization algorithm in mobile sensor networks CTCOMSN is presented. Simulation is done in obstacle coverage phenomena and circumstances that static nodes can not complete monitoring, and the results show that CTCOMSN is efficient in improving coverage quality and prolonging network lifetime. CTCOMSN can also be used in normal coverage.(6) For multiple target monitoring, several targets are monitored by a group of sensor nodes, while several sensor nodes are in charge of many targets. Relationship between them is complex, and there must be some association among nodes and targets. According to that association, multiple targets association coverage algorithm MTCAC is designed using association rule dining technology. Network lifetime is prolonged by transition of the work state of sensor nodes groups. Compared with advanced PEAS, MTCAC shows better coverage quality and longer lifetime. MTCAC is the first coverage algorithm in the view of association among nodes and targets, which illuminates other association research in WSN.(7) The location of sensor nodes can influence the coverage quality. To adapt sensor nodes to deployment of the physical characters, a novel coverage algorithm NGECA is presented. It is also the first algorithm in view of physical characters inside the monitored area, which is significant in other WSN research. Simulations have studied on influence of space grid and physical gradients, and whose results show that NGECA can provide high coverage quality with a few sensor nodes, while network resource is saved at the same time and robust and stability of WSN is increased.(8) In order to satisfy the real-time requirement in long-time environmental applications, a real-time NGECA is presented. Simulation results show that NGECA can quickly provide the physical information if the character varied in monitored area. It is a would-be 3D algorithm, and gives some ideas to the future real-time coverage research.更多还原