【作者】 凡高娟；

【导师】 王汝传；

【作者基本信息】 南京邮电大学 ， 信息网络， 2010， 博士

【摘要】 无线传感网络(Wireless Sensor Networks, WSNs)在新一代网络中具有关键性角色,作为一种新的计算模式,推动科技发展和社会进步,已成为国际竞争的焦点和制高点,关系到国家政治、经济和社会的安全。在无线传感器网络应用中,覆盖反映了网络对物理世界的监测能力,常作为描述无线传感器网络监测服务质量(Quality of Service, QoS)的标准,是无线传感器网络应用的基础。由于应用规模大,能量、计算能力和通信能力受限等特点,往往在监测区域内对传感器节点进行大规模、高密度部署,这就造成部署代价与部署质量、节点能量受限与网络生存时间之间的矛盾。覆盖控制通过对节点感知能力的建模,结合具体应用要求,在监测区域内部署传感器节点,使其能有效的获取监测区域的信息;在不影响网络覆盖性能的前提下,合理调度节点状态以减少网络中活跃节点数量,延长网络生存时间;最终使传感器网络的各种资源得到有效利用,网络监控服务质量得到改善。本文深入研究无线传感器网络覆盖控制技术,开展无线传感器网络节点部署和节点调度问题研究。本文主要工作和贡献如下:1)提出两种覆盖保证的节点随机部署策略:平均面积覆盖部署策略和边界辅助部署策略,以解决节点部署中边界效应造成的覆盖过高估计问题,满足实际部署中覆盖质量要求。部署质量和部署代价是无线传感器网络相矛盾的两个方面。网络的成功应用依赖于节点部署中所达到的覆盖质量,即用较少的节点达到所期望的覆盖要求。通常用渐近性分析来求出满足覆盖要求的最少节点数。然而,由于存在覆盖过高估计问题,这种渐近性分析在实际的部署中难以达到部署要求。针对该问题,提出两种覆盖保证的节点部署策略:平均面积覆盖部署策略ECD(Expected-area Coverage Deployment)和边界辅助部署策略BOAD(BOundary Assistant Deployment)。从理论上对两种部署策略进行分析,给出满足部署要求节点个数的下限值。通过理论分析与仿真验证两种部署策略正确性和有效性。仿真结果表明,两种部署策略有效的缓解了覆盖过高估计问题。在目标检测应用中验证了两种部署策略的优势,并把两种部署策略扩展到一般监测区域和节点概率感知模型中。2)提出一种距离辅助的节点覆盖冗余判别模型,节点只需要获得与邻居的距离信息,就可以精确判别被邻居节点覆盖的程度。并在此基础上提出非均匀分布的无线传感器节点调度机制。针对传统依赖精确位置信息的复杂计算和无位置信息部署受限等弊端,从理论上对节点部署方式进行分析。在节点位置信息未知情况下,提出距离辅助的节点覆盖判别模型DANCI(Distance-Assistant Node Coverage Identification model),通过邻居节点到本节点的距离信息,判别该节点被邻居节点覆盖的程度。并在此基础上提出一种非均匀分布下的无线传感器网络节点调度机制NDNS(Non-uniform Distribution Node Scheduling),该机制利用距离辅助的节点覆盖判别模型,对节点覆盖冗余进行判别,适应于任意分布下的网络部署方式。仿真结果表明,在节点随机部署情况下,计算所得的覆盖冗余与节点位置已知情况下的最大误差仅为6.3960%。在保证网络覆盖的前提下,有效的延长了网络的生存时间。3)提出一种基于容忍覆盖区域的节点调度算法,解决位置信息未知条件下,由于不均等休眠带来的调度洞问题。位置信息未知的节点调度算法以节点的感知区域覆盖为调度目标,导致处于边界区域的节点由于没有太多机会进入休眠状态而先死亡,进而引起死亡节点向监测中心扩散现象;或者监测区域中心某些孤立节点由于没有太多机会休眠先死亡,引起死亡节点由监测区域中心向外部扩散现象,我们称这种现象为“不均等休眠”带来的调度洞问题。针对该问题,从理论上对节点覆盖模型进行分析,提出容忍覆盖区域的概念,并在此基础上设计一种基于容忍覆盖区域的节点调度算法。仿真结果表明,基于容忍覆盖区域的节点调度算法能够缓解“不均等休眠”现象,解决了由边界节点或孤立节点过早死亡造成的能量消耗不平衡问题,有效的延长了网络的生存时间。4)提出在实际应用场景中入侵目标至少被k个节点立刻检测到的理论模型,分析外界环境等因素对节点感知能力的影响,得出目标入侵监测区域时被检测到的概率。目标检测和环境检测是传感器网络的重要应用之一。节点的监测质量可以通过入侵目标穿越监测区域时被检测的概率来衡量,节点检测的概率为监测质量和决策提供基础。然而,在实际的网络部署和应用场景中,节点的感知能力通常受到阴影衰落及路径损失的影响。在此情况下,提出在实际应用场景中入侵目标至少被k个节点检测到的理论模型,并详细分析了节点在阴影衰落与理想感知检测两种情况下的检测概率,进而说明理想的感知检测与实际应用中检测概率的明显差异。通过该理论分析模型可以有效的估计出节点在具体的应用场景下的监测性能。更多还原

【Abstract】 Wireless Sensor Networks (WSNs) play an important role in next generation networks. As an emerging technology with great potential to drive the development of science and technology, promote the advancement of society, and impact national politics and security, WSNs have become a primary focus of international competitors. In wireless sensor networks applications, network coverage reflects an ability of monitoring the physical world, often as a description to monitor standards of the quality of service (Quality of Service, QoS). As the application characteristic for large scale, the energy constraints, with limited computing ability and communication ability. The sensor nodes are often deployed with large-scale, high-density in monitored region, which brings conflicts between deployment costs and deployment quality, the energy constraints of nodes and network lifetime. Coverage control by modeling the sensing ability of sensor nodes, combined with application requirements, effectively obtain information in monitoring area by deployment of sensor nodes. And without sacrificing system original performance by putting some sensor nodes into active state for the sensing and communication tasks while other sensor nodes remain low-power sleep state. The network lifetime is prolonged by efficiently using a variety of resources. Finally, the QoS is improved in monitored region.In this dissertation, we mainly focus on coverage control problems in wireless sensor networks. The outline of this dissertation is described as follows:1) Two coverage-guaranteed sensor node deployment strategies for wireless sensor networks are proposed to overcome coverage overestimation problem induced by border effects. Eg. Expected-area Coverage Deployment (ECD) and BOundaryAssistant Deployment (BOAD), and then coverage quality requirements can be satisfied in real deployment. Deployment quality and cost are two conflicting aspects in wireless sensor networks. Their successful applications depend considerably on the deployment quality that uses the minimum number of sensors to achieve a desired coverage. Currently, the number of sensors required to meet the desired coverage is based on asymptotic analysis, which cannot meet deployment quality due to coverage overestimation in real applications. To overcome this problem, we propose two deployment strategies, namely, the expected-area coverage deployment strategy and boundary assistant deployment strategy. The deployment quality of the two strategies is analyzed mathematically. Under the analysis, a lower bound on the number of deployed sensor nodes is given to satisfy the desired deployment quality. We justify the correctness of our analysis through rigorous proof, and validate the effectiveness of the two strategies through extensive simulation experiments. The simulation results show that both strategies alleviate the coverage overestimation significantly. In addition, we also evaluate two proposed strategies in the context of target detection application. The comparison results demonstrate that if the target appears at the boundary of monitored region in a given random deployment, the average intrusion distance of BOAD is considerably shorter than that of ECD with the same desired deployment quality. In contrast, ECD has better performance in terms of the average intrusion distance when the invasion of intruder is from the inside of monitored region. The general monitored region and probabilistic sensing model are extended for two deployment strategies.2) A distance-assistant node coverage identification model is proposed by using distance information between the node and its neighbors to identify nodes coverage without using any location information. The coverage degree can accurately determined by the neighbor nodes. Based on this model, a NDNS (Non-uniform Distribution NodeScheduling) is proposed which satisfies with different random distribution. Aiming at the defect that high computational complexity of exact location information and the distribution limitation of location information-free in traditional schemes, the node distribution is analyzed theoretically. A distance-assistant node coverage identification model(DANCI) is proposed which adopts distance information between the node and its neighbors to identify nodes coverage without using any location information. A node scheduling scheme NDNS based on DANCI is proposed which satisfies with different random distribution. Theory analysis and simulation results are presented to evaluate the proposed DANCI model. It is shown that in randomly deployed sensor networks, the maximum coverage error is 6.396 0% between DANCI model and location information-aware strategy. The numerical experiments results illustrate that the longer network lifetime is achieved in preserving networks coverage.3) A node scheduling scheme based on tolerable coverage area is proposed. The node scheduling problem conduced by the inequality sleeping in location-free conditions is alleviated by proposed algorithm. Traditional methods of node scheduling without location information are aiming at node sensing area coverage. It leads to nodes in the border of monitored region or some nodes without neighbors in monitored region die first due to having no chance to enter sleep state, and then the death spreads to the central region. We call this phenomenon as inequality sleep problems. To address this problem, from the theoretical analysis of the sensor node coverage model, we propose the concept of tolerable coverage area, and a node scheduling scheme based on tolerable coverage area. Simulation results demonstrate that the proposed method not only alleviates the inequality sleep problems, but also prolongs network lifetime.4) An intrusion detection model is suggested, in which the intruder can be detected immediately by at least k sensor nodes under practical considerations. The problem that the sensing capabilities of sensors are affected by environmental factors in real deployment is investigated. The detection probability by at least k sensors under practical considerations is studied. Target detection and field surveillance are among the most prominent applications of wireless sensor networks. The quality of detection achieved by a sensor network can be quantified by evaluating the probability of detecting a mobile target crossing a sensing field. Detection probability of sensor nodes has been studied in sensor networks for many purposes such as quality of service and decision-making. However, the sensing capabilities of sensors are affected by environmental factors in real deployment. The problem of detecting probability in a log-normal shadow fading environment is investigated. It presents an analytic method to evaluate the detection probability by at least k sensors under practical considerations. Furthermore, we also shows that shadow fading makes significant influence in detection probability compared to unit disk sensing model through extensive simulation experiments.更多还原