一种基于约束策略的无线传感器网络定位算法*

节点定位技术是无线传感器网络的关键支撑技术之一,对于无线传感器网络的基本理论方法和应用研究都具有重要意义。在深入研究分析距离无关定位算法的基础上,提出了基于约束策略的无线传感器网络定位算法。该算法无须测距,采用跳数估计节点间距离,并针对未知节点到锚节点距离计算中的不足,对锚节点的平均每跳距离作了修正;在估计未知节点坐标时,根据该未知节点通信范围内的锚节点对其所在位置进行约束。仿真结果表明,该算法具有较好的性能,比已有算法的定位精度有所提高。     

一种适用于无线传感器网络的健壮的有序定位算法

提出了一种分布式无线传感器网络有序定位算法,利用邻居节点间的测量距离和两跳邻居的坐标信息来对网络中的节点进行定位,在测量距离误差较大的情况下仍然能够比较理想地估算出节点的地理位置.先对约束条件比较多的节点进行定位,这样该节点的定位精度就比较高,一个节点得到坐标后又会引入若干约束条件,这些条件又作为定位其他节点的约束,就这样一直定位下去.实验证明此种优化手段可以显著改善定位精度.详细分析了该定位算法中采用的各种技术,并针对邻居数目和测量距离精度做了很多实验来研究其对定位结果的影响.     

基于误差修正的DV-Hop传感器节点定位算法

传统的DV-Hop传感节点定位算法,估计未知节点与各锚节点之间距离是用跳段距离代替直线距离.在实际网络定位环境中,未知节点和锚节点之间多数是折线连接.当平均每跳距离的估计值与实际值的偏差较大时,未知节点到锚节点之间估计距离与实际距离之间的误差会增大.为解决上述问题,提出一种粒子群优化算法修正DV-Hop算法定位误差的传感器节点定位方法.采用DV-Hop算法估计待测节点和锚节点之间距离,通过三边测量法确定节点的位置,并将传感器节点定位问题转换成一个多约束优化问题,最后通过粒子群优化算法对定位误差进行修正,并通过仿真对其性能进行测试.仿真结果表明,相对传统DV-Hop算法可大幅度提高传感器节点定位精度,符合无线传感器网络定位需求,具有较好的应用价值.     

基于无线网络节点相似度的室内定位算法

针对RSSI测距定位技术在室内环境中受到的随机干扰大,干扰变化情况不确定,定位结果误差较大的问题,提出了一种基于无线传感器网络节点相似度的室内定位算法。首先通过无线传感器网络的连通特性确定位于网络盲节点周围的几个信标节点,利用数据拟合的方法确定节点工作电压对RSSI距离测量的影响,然后根据拟合结果对不同工作电压下测得的RSSI值进行修正,利用修正后得到的RSSI值计算得到网络节点之间的相似度值,并利用该相似度值作为距离测量和定位结果的自校正系数,对定位结果进行修正,从而获得精度较高的定位效果。实验结果表明该定位算法具有较高的定位精度。     

基于无线传感器网络的同时定位与跟踪

研究无线传感器网络在位置信息不确定时,同时定位无线传感器网络节点并跟踪移动目标。利用RSSI测量节点对之间的距离,多维定标技术根据距离矩阵完成传感器网络的初始定位。估计与更新阶段提出了压缩EKF滤波确定传感器节点位置和目标位置。仿真结果显示:算法在较低的网络覆盖率下有较高的定位和跟踪准确度,在初始定位误差为5m时,节点和跟踪误差均小于3m,特别是在长距离的跟踪任务中有很好的精度和实时性。     

基于RSSI测距的传感器网络定位算法研究

基于RSSI(Received Signal Strength Indicator)的测距技术根据理论或经验信号传播模型将传播损耗转化为距离.在实际应用环境中,由于多径、绕射、障碍物等因素,无线电传播路径损耗使得定位过程中产生距离误差.通过对二维空间定位过程中产生距离误差区域进行分析,提出了基于RSSI的新的定位算法ERSS.该算法计算简单,定位过程中节点间不增加通信开销,无需硬件扩展.仿真实验表明,该算法较普通的基于RSSI的测距方法有了明显的改进,提高了距离估计的精度,适合在通信开销小、硬件要求低的传感器网络节点上应用.     

一种分布式无线传感器网络节点定位算法

基于接收信号强度(RSSl)的无线传感器网络节点定位算法由于无需额外测距硬件的支持而受到广泛的应用.但无线信号传输受到环境的影响,使得基于RSSI的定位算法存在较大的定位误差.针对上述不足提出了一种基于概率的无线传感器网络节点定位方法.首先根据信标节点发送信号强度与未知节点接收信号强度差来计算未知节点到其一跳范围内所有信标节点的估计距离,然后根据这些估计距离,通过计算未知节点在某个位置概率密度函数的最大值来确定未知节点自身的位置坐标.通过仿真表明,该方法在较低的信标节点密度和节点通信半径条件下,具有较高的节点定位精度和定位覆盖率.     

基于遗传模拟退火算法的无线传感器网络定位算法

无线传感器网络定位问题本质上是一个基于不同的距离或路径测量方法的优化问题.第一次提出了基于遗传模拟退火算法的无线传感器网络定位算法GASA-Hop,它是将遗传模拟退火算法作为DV-Hop的后期优化.其中,DV-Hop用来估计未知节点与锚节点的测量距离,GASA用来最小化与DV-Hop相关的适应度函数.仿真结果表明,本算法定位精度高、条件简单,比较适合无线传感器网络的节点定位.     

无线传感网络节点定位中的导标动态移动策略

在基于移动导标的无线传感器网络节点定位中,导标的移动路径将直接影响到节点定位的效率和精度.考虑到无线传感器网络节点通信能力和计算能力有限性,着重研究一种基于网络局部拓扑信息的导标移动虚拟力修正模型.模型中,导标对邻居节点进行筛选,并进一步收集有效节点与导标的距离、有效节点拥有邻居节点数目等参数,在此基础上建立移动导标虚拟引力模型,该模型能使导标的移动具有较强的自适应性.更进一步,针对节点虚拟引力基本模型存在的导标无引力情况和冗余遍历情况,通过加入接近因子及已定位区域边界曲线拟合的方法,对基本模型进行修正.仿真实验表明,提出的路径规划算法与典型的传统路径算法相比较,导标遍历网络的路径长度缩短了20%～30%,较大程度节约了定位过程中导标节点遍历网络所需的能量开销.     

无线传感器网络中基于能量检测的定位研究

在许多无线传感器网络的应用中,节点定位是其关键技术之一,而定位精度的高低在很大程度上取决于距离测量的精度。针对射频信号受各种干扰影响而出现测距不精确的问题,该文对能量检测采用改进的限幅平均滤波方法,校正测量的节点间点到点的距离,并优选信标节点,最后用加权最小二乘法进行定位。实验及仿真结果表明,本定位算法可以提高测距精度...     

基于移动信标的无线传感器网络节点定位技术综述

无线传感器网络已在许多领域有着广泛的应用,网络中的节点定位是为诸多应用提供了基本信息,基于移动信标的无线传感器网络节点定位是重要的研究领域之一。介绍了基于移动信标的无线传感器网络节点定位的基本原理和评价标准,对该技术的未来进行了展望。     

移动节点路径规划和几何限制的无线传感器网络定位算法

提出一种利用移动节点的无需测距的无线传感器网络定位算法.该算法中,移动节点以垂直路径两次穿过未知节点通信半径范围,从而获得通信区域边界附近多个航标位置;航标点连线的中垂线形成几何限制区域,该区域中心即为未知节点的估计位置.与其它基于几何限制区域的算法相比,本算法计算复杂度低、定位精度高.仿真实验结果显示本算法相比于其它算法,定位精度提高10％～40％不等.     

使用移动锚节点的增量式自定位算法研究

无线传感网节点自定位技术是许多相关应用的前提和基础,目前已提出多种定位算法,但大多用于静态无线传感网。针对使用移动锚节点定位场景提出一种基于测距的算法PMAIL（PSO-based Mobile Anchor Incremental Localization）,将节点精度分级,选择高等级参考节点进行增量式定位,同时使用粒子群算法（PSO）求解加权误差方程,得到最优位置估计。算法不局限于特定测距方式,锚节点可同时支持常见的移动sink数据收集和网络管理等功能。仿真表明算法有较高的网络覆盖率,精度提高接近9%。     

基于声誉机制的传感器网络节点安全定位算法

针对传感器网络中的节点安全定位问题,提出一种基于声誉机制的安全定位算法。算法引入信标节点的相互监督机制,使用Beta分布来表示更新和整合信标节点的声誉值,通过簇头节点来收集并判断哪些信标节点是信誉值高的可信节点,并应用声誉模型来排除提供错误定位信息的恶意信标节点,提高了恶意信标节点的检测率,减少了定位误差,提高了定位系统的鲁棒性;通过仿真实验详细分析了定位算法的有效性和鲁棒性,所提算法适用于分布式传感器网络节点实现自身定位,且在定位精确性和安全性方面都有很大提升。     

一种改进的自适应环境定位方法

随着无线传感器网络技术的不断进步成熟, 其应用将会越来越广泛, 实用化的节点定位技术研究对于传感器网络技术性能提高和实用性保证有重要的理论意义和应用价值. 提出一种环境自适应的RSSI户内定位方法. 该方法在自修正环境路径损耗指数的基础上, 由初次定位获取各信标节点定位误差, 以误差为加权质心算法的权值计算盲节点的坐标, 并对坐标误差进行补偿. 在NI无线传感器网络(WSN)平台上进行的验证结果表明, 该方法具有更高的定位测量精度和稳定性.     

WSN中信标节点移动情况下的定位方法研究

定位技术是无线传感器网络的基础理论和关键技术之一.在实际应用中,一些信标节点在部署以后其位置由于各种原因而会发生漂移,由此就使得依赖于这些信标节点来定位的其他节点将无法准确定位,或者产生很大定位误差.本文针对信标节点产生漂移情况下的节点定位问题,提出了一种基于区域划分的信标节点移动检测算法(AD-BMD),在区域划分的基础上,通过两次判定过程将移动的信标节点检测出来,并为每个信标设置了坐标可信度值.为合理利用移动的和未移动的信标节点,提出一种基于信标移动检测的信标择优定位算法(BMD-BOS),通过合理选择信标节点来对其它未知节点进行定位.实验结果证明,AD-BMD算法具有很高的检测正确率和较低错误率,检测效果好于LB和SSV方法.BMD-BOS算法的定位精度远远高于没有进行移动信标检测的定位算法(N-BMD)和将移动信标丢弃的定位算法(D-BMD).     

Energy-efficient multipath routing in wireless sensor network considering wireless interference

Due to the energy and resource constraints of a wireless sensor node in a wireless sensor network (WSN), design of energy-efficient multipath routing protocols is a crucial concern for WSN applications. To provide high-quality monitoring information, many WSN applications require high-rate data transmission. Multipath routing protocols are often used to increase the network transmission rate and throughput. Although large-scale WSN can be supported by high bandwidth backbone network, the WSN remains the bottleneck due to resource constraints of wireless sensors and the effects of wireless interference. In this paper, we propose a multipath energy-efficient routing protocol for WSN that considers wireless interference. In the proposed routing protocol, nodes in the interference zone of the discovered path are marked and not allowed to take part in the subsequent routing process. In this way, the quality of wireless communication is improved because the effects of wireless interference can be reduced as much as possible. The network load is distributed on multiple paths instead of concentrating on only one path, and node energy cost is more balanced for the entire wireless network. The routing protocol is simulated in NS2 software. Simulation result shows that the proposed routing protocol achieves lower energy cost and longer network lifetime than that in the literature.     

Enhanced Security with Improved Defensive Routing Mechanism in Wireless Sensor Networks

In recent scenario of Wireless Sensor Networks (WSNs), there are many application developed for handling sensitive and private data such as military information, surveillance data, tracking, etc. Hence, the sensor nodes of WSNs are distributed in an intimidating region, which is non-rigid to attacks. The recent research domains of WSN deal with models to handle the WSN communications against malicious attacks and threats. In traditional models, the solution has been made for defending the networks, only to specific attacks. However, in real-time applications, the kind of attack that is launched by the adversary is not known. Additionally, on developing a security mechanism for WSN, the resource constraints of sensor nodes are also to be considered. With that note, this paper presents an Enhanced Security Model with Improved Defensive Routing Mechanism (IDRM) for defending the sensor network from various attacks. Moreover, for efficient model design, the work includes the part of feature evaluation of some general attacks of WSNs. The IDRM also includes determination of optimal secure paths and Node security for secure routing operations. The performance of the proposed model is evaluated with respect to several factors; it is found that the model has achieved better security levels and is efficient than other existing models in WSN communications. It is proven that the proposed IDRM produces 74% of PDR in average and a minimized packet drop of 38% when comparing with the existing works.     

Efficient and accurate sensor network localization

Wireless sensor networks (WSN) have great potential in ubiquitous computing. However, the severe resource constraints of WSN rule out the use of many existing networking protocols and require careful design of systems that prioritizes energy conservation over performance optimization. A key infrastructural problem in WSN is localization—the problem of determining the geographical locations of nodes. WSN typically have some nodes called seeds that know their locations using global positioning systems or other means. Non-seed nodes compute their locations by exchanging messages with nodes within their radio range. Several algorithms have been proposed for localization in different scenarios. Algorithms have been designed for networks in which each node has ranging capabilities, i.e., can estimate distances to its neighbours. Other algorithms have been proposed for networks in which no node has such capabilities. Some algorithms only work when nodes are static. Some other algorithms are designed specifically for networks in which all nodes are mobile. We propose a very general, fully distributed localization algorithm called range-based Monte Carlo boxed (RMCB) for WSN. RMCB allows nodes to be static or mobile and that can work with nodes that can perform ranging as well as with nodes that lack ranging capabilities. RMCB uses a small fraction of seeds. It makes use of the received signal strength measurements that are available from the sensor hardware. We use RMCB to investigate the question: “When does range-based localization work better than range-free localization?” We demonstrate using empirical signal strength data from sensor hardware (Texas Instruments EZ430-RF2500) and simulations that RMCB outperforms a very good range-free algorithm called weighted Monte Carlo localization (WMCL) in terms of localization error in a number of scenarios and has a similar computational complexity to WMCL. We also implement WMCL and RMCB on sensor hardware and demonstrate that it outperforms WMCL. The performance of RMCB depends critically on the quality of range estimation. We describe the limitations of our range estimation approach and provide guidelines on when range-based localization is preferable.     

LAD: Localization anomaly detection for wireless sensor networks

In wireless sensor networks (WSNs), sensors’ locations play a critical role in many applications. Having a GPS receiver on every sensor node is costly. In the past, a number of location discovery (localization) schemes have been proposed. Most of these schemes share a common feature: they use some special nodes, called beacon nodes, which are assumed to know their own locations (e.g., through GPS receivers or manual configuration). Other sensors discover their locations based on the reference information provided by these beacon nodes. Most of the beacon-based localization schemes assume a benign environment, where all beacon nodes are supposed to provide correct reference information. However, when the sensor networks are deployed in a hostile environment, where beacon nodes can be compromised, such an assumption does not hold anymore. In this paper, we propose a general scheme to detect localization anomalies that are caused by adversaries. Our scheme is independent from the localization schemes. We formulate the problem as an anomaly intrusion detection problem, and we propose a number of ways to detect localization anomalies. We have conducted simulations to evaluate the performance of our scheme, including the false positive rates, the detection rates, and the resilience to node compromises.