摘要
无线传感器网络是由大量密集部署在目标区域的自治节点构成的一种自组织网络应用系统。它综合了无线通信技术、嵌入式技术、传感器技术、分布式信息处理技术等,是国内外公认的可以推动信息领域进入崭新发展阶段的新兴技术,在军事侦察、医疗监护、环境监测、交通管理、反恐防灾、智能家居等领域具有广泛的应用前景。
无线传感器网络是由一组具有一定移动性的传感器节点组成的,是集数据采集、处理和传输于一体的分布式网络。覆盖问题是影响其工作效果的一个重要因素。在网络初始分布时,由于通常节点个数较多,采用人工的方法进行逐一放置显然不现实,取而代之的是常用的随机布撒的方式。但是,这样很难保证覆盖结果具有足够的均匀性,也就影响着网络的运行效率。因此,节点需要依据所设计的移动覆盖算法进行适当的位置调整。
良好的移动覆盖算法应该使各移动节点依据且只依据周围临近节点的分布情况动态地调整自身位置,从而使整个网络覆盖趋向均匀;同时,在网络某些节点失效时能够较快地响应网络拓扑的变化。
本文的目的是研究无线传感器网络的协作覆盖问题,设计与验证有效的解决方法。
本文基于图论与机器人运动学,针对无线传感器网络分布式协作与多跳通信特点,提出了一种动态模型。该模型用Delaunay三角剖分和Voronoi图描述相邻节点的几何关系。模型中,每个节点的行为只与其一跳相邻节点和其所处环境有关。该模型为无线传感器网络中各种任务的完成提供了一个解决方案,例如网络的自组织覆盖,网络数据查询路由,相邻节点的信息共享。
为解决无线传感器网络连接与覆盖问题,本文提出了两种自组织算法。第一种是改进的虚拟力算法,它将虚拟力与粒子群相结合,先用改进型粒子群算法对这两个系数寻优。然后利用寻优结果计算出虚拟力并部署节点。仿真表明在引力与斥力共同作用下,传感器网络能够兼顾网络节点部署的快速性与最终的覆盖率。第二种方法是基于生物竞争的自组织覆盖算法,该方法受自然界生物间争夺资源的启示,将静态传感器定义为“强势个体”,将可移动传感器定义为“弱势个体”,将每一个传感器节点的有效覆盖面积定义为该节点所获取的“自然资源”。仿真证明,生物竞争法能够很好的布置兼有动静态节点的网络。
为了验证自组织方法的有效性,本文提出几种直观量化的自组织评价方案。利用Delaunay三角剖分评价节点实体和他们的关系以及结点之间的信息传递和融合;利用Voronoi图进行评价节点覆盖的区域;同时,对整个自组织过程,引用自组织度的概念对其分布效果进行定量分析。
针对虚拟力自组织法中节点运动轨迹抖动、耗能大的问题,我们结合生物神经分流模型,提出了一种基于分流模型自组织路径平滑控制策略。由于分流模型的输出是稳定、光滑且有界的,故将分流模型与虚拟力控制相结合,有效地平滑了传感器网络自组织过程中节点的轨迹。
本文还提出了同步和异步两种编队策略,仿真结果表明,异步法都能有效的使网络编队移动,顺利的到达新的目标覆盖区域;同步法能够完好的保持网络原有的结构,极大的减少节点间相互定位的损耗,有效的保持了相互通信,有极好的鲁棒性。
Wireless sensor network is a self-organizing network application system composed of a large number of nodes in the target region. It combines the embedded technology, wireless communication technology, sensor technology and distributed information processing technology. It is an emerging technology recognized at home and abroad which can promote the field of information into the new stage of development. Wireless sensor network has wide application prospects in military reconnaissance, environmental monitoring, medical care, traffic management, intelligent home, anti-terrorism, disaster preparation and other areas.
Wireless sensor network is composed of a group of mobile sensor nodes, and is a distributed network combining data acquisition, processing and transmission. The issue of covering is an important factor for its work performance. For the initial network distribution, because the large number of nodes, the method of random scraps is commonly used replacing the artificial method However, it is very difficult to guarantee sufficient coverage of the results and thus affects the operating efficiency of the network. Therefore, the node location is required to be adjusted by the mobile coverage algorithm.
Good mobile coverage algorithm should make the mobile nodes adjust their own positi- ons only on the basis of the distribution of nodes nearby dynamically, tending to evenly cover the entire network. At the same time, for the failure of some nodes the network is able to qui- ckly respond to network topology changes.
The purpose of this paper is to study collaboration coverage problem of wireless sensor network collaboration, design and verify the effective solutions.
Based on graph theory and robot kinematics, this paper proposes a dynamic model for distributed collaboration and multi-hop communication of wireless sensor networks. The model uses Delaunay triangulation and Voronoi diagram to describe the geometric relationship between adjacent nodes. In the model, each node acts only according to its neighboring nodes and environment. The model provides a solution for wireless sensor networks in a variety of tasks, such as self-organizing network coverage, network data query routing and information sharing of adjacent nodes.
To solve the problem of connection and covering of wireless sensor network, this paper presents two self-organizing algorithms. The first one is the improved virtual force algorithm combining virtual force and particle swarm. It uses the improved particle swarm optimization to seek two factors and then calculates the virtual force and deploys the nodes. Simulation shows that in the combined effect of gravitational force and repulsion force, the sensor net- work can bring rapid deployment of network nodes and final coverage. The second method is self-organizing covering algorithm based on biology competition. Inspired by the resources fights in the natural world, this method defines static sensors as‘strong individuals’and mo- bile sensors as‘disadvantaged individuals’, and the effective coverage of each sensor node as‘natural resources’. Simulation proves that the biology competition method can well layout the network with both static and dynamic nodes.
To verify the effectiveness of self-organizing method, this paper proposes several intuitive and quantitative assessment methods. Delaunay triangulation is used to evaluate the node entities and their relationships as well as the message transmission and integration. Voronoi diagram is used to evaluate the coverage of the nodes. At the same time, for the whole process of self-organization, the concept of self-organization degree is introduced to analyze its distribution effect quantitatively.
To solve the problem of large energy consumption and trajectory jitter, this paper proposes self-organization diverted path smoothing control strategy based on streaming model combining with neural biology. Because the output of streaming model is stable, smooth and bounded, the combination of streaming model and virtual force control can effectively smooth the trajectory in the self-organization of sensor network.
This paper also proposes synchronous and asynchronous formation strategies. Simulation results show that the asynchronous method can effectively make the network move in formation and get to the new target coverage area. And the synchronous method can maintain the original structure of the network and greatly reduce the loss in the node location, which effectively maintains the mutual communication and has excellent robustness.
引文
1.王蓓,孟伟民.无线传感器网络在信息战中的运用探讨[J].中国科技信息, 2005(23): 44.
2.牟连佳,牟连泳.无线传感网络及其在工业领域应用研究[J].工业控制计算机, 2005, 18(1):3-5.
3.冯友兵,张荣标,沈敏等.面向精确灌溉的无线传感器网络构建[J].农业机械学报, 2009, 40 (1):56-59.
4.马碧春.无线传感器网络在医疗行业的应用展望[J].中国医院管理, 2006, 26(10):73.
5.田欣,胡莲.基于无线传感网的城市应急指挥系统关键技术初探[J].经济师, 2006 (6): 53-54.
6.饶云华,代莉,赵存成等.基于无线传感器网络的环境监测系统[J].武汉大学学报(理学版), 2006, 52(3):345-348.
7.李光辉,赵军,王智.基于无线传感器网络的森林火灾监测预警系统[J].传感技术学报, 2006, 19(6):2761-2764.
8.李莹莹,刘云辉,樊玮虹等.基于移动通信网络的机器人遥操作[J].通信学报, 2006, 27 (5):52-59.
9.张志海,沈连丰,叶芝慧等.一种基于Zigbee的无线传感器网络教学实验系统[J].电气电子教学学报, 2006, 28(5):76-80.
10.高爱国,马海武,王艳阳等.无线传感器网络在智能家居中的应用[J].中国西部科技, 2009, 08(03):18-19.
11. Leonard J, Durrant-Whyte H. Mobile robot localization by tracking geometric beacons[J]. IEEE Transactions on Robotics and Automation, 1991, 7(3):376-382.
12. Guivant J, Nebot E. Optimization of the simultaneous localization and map-building alg- orithm for real-time implementation[J]. IEEE Transactions on Robotics and Automation, 2001, 17(3):242-257.
13. Fox D, Burgard W, Kruppa H. A probabilistic approach to collaborative multi-robot loca- lization[J]. Autonomous Robots, 2000, 8(3):325-344.
14. Thrun S, Fox D, Burgard W, et al. Robust Monte Carlo localization for mobile robots art- ificial intelligence[J]. Artificial Intelligence, 2000, 128(1):99-141.
15. Zou Y, Chakrabarty K. Sensor deployment and target localization based on virtual forces[J]. ACM Transactions on Embedded Computing Systems, 2004, 3(1):61-91.
16.王德胜.无线传感器网络节能策略研究[D]:[硕士学位论文].无锡:江南大学信控学院,2008.
17.汪学清,杨永田.一种基于虚拟菱形网格的传感器节点布置算法[J].计算机应用, 2006, 26(7):1554-1556.
18.吴彤.自组织方法论研究[M].北京:清华大学出版社,2001.15-69
19.陈志,王汝传,孙力娟.无线传感器网络的自组织机制研究[J].电子学报, 2007,5(35):854-857.
20. S.Grossberg. Nonlinear neural networks:Principles,mechanism,and architectures[J]. Ne- ural Networks(S0893-6080),1988,1(1):17-61.
21. Intanagonwiwat C, Govindan R, Estrin D. Directed diffusion for wireless sensor netwo- rking[J]. IEEE/ACM Transactions on Networking, 2003, 11(1):2-16.
22.秦宁宁,盖祎,张林,蒋敏峰,徐保国等. Voronoi图在无线传感器网络栅栏覆盖中的应用研究[J].计算机应用研究,2008,25(3):863-865.
23.王雷,沈昊为,林亚平.传感器网络中基于Voronoi网格的快速覆盖判定算法[J].系统仿真学报,2008,20(14):3585-3563.
24.公维宾,常义林,沈中,张颖.传感器网络中基于最小覆盖重叠的移动部署[J].系统仿真学报,2008,20(13):3604-3609.
25.黄艳,臧传治,于海斌.传感器网络中无线电干涉定位系统的多径误差分析[J].控制与决策,2009,24(2):231-235.
26.信俊昌,王国仁.传感器网络中自顶向下的轮廓监控算法[J].控制与决策,2009, 24(2): 191-195.
27.王季荣,刘卫东.基于节点活性的移动传感器网络传输算法[J].清华大学学报(自然科学版),2008,48(10):1663-1666.
28.于振华,刘宇,蔡远利.基于扩展合同网协议的无线传感器网络协作方法研究[J].控制与决策,2009, 24(1):61-65.
29.郑巍,刘三阳,寇晓丽.基于模块度的无线传感器网络分簇算法[J].系统仿真学报,2009,21(1):65-67.
30.刘丽萍,李桂丹,王智,孙雨耕.能量有效的无线传感器网络协同覆盖[J].计算机工程与应用, 2008,44(34):31-34.
31.陶丹,马华东,刘亮.视频传感器网络中路径覆盖增强算法研究[J].电子学报,36(7): 191-196.
32.胡劲文,梁彦,潘泉.无线传感器网络部分覆盖算法及连通性研究[J].计算机工程与应用, 2008,44(34):24-27.
33.何弢,徐昌庆.无线传感器网络的动态部分覆盖算法[J].信息技术,2008,(9):48-51.
34.向满天,史浩山,李立宏.无线传感器网络的连通覆盖临界条件分析[J].传感技术学报,2008,21(11):1887-1891.
35. Nurcan Tezcan, Wenye Wang. Self-orienting wireless multimedia sensor networks for occlusion-free viewpoints[J].Computer Networks, 2008(52):2558-2567.
36. Luo Wusheng, Zhai Yongping, Lu Qin. Study on Wireless Multimedia Sensor Networks[J]. Journal of Electronics & Information Technology, 2008,30( 6):1511-1515.
37. Tao Dan, Ma Huadong, Liu Liang . A Virtual Potential Field Based Coverage-Enhancing Algorithm for Directional Sensor Networks[J].Journal of Software,2008,18(6):1152-1163.
38. You-Chiun Wang, Yu-Chee Tseng. Distributed Deployment Schemes for Mobile Wireless Sensor Networks to Ensure Multilevel Coverage[J]. IEEE Transactions on Parallel and Distributed Systems, 2008,19(9):1280-1294.
39. Sohrabi K, Gao J, Ailawadhi V and Pottie GJ. Protocols for Self-organization of a Wireless Sensor Network[J]. IEEE Personal Communications, 2000,7(5):16-27.
40. Tao Dan, Ma Huadong, Liu Liang. Study on Path Coverage Enhancement Algorithm for Video Sensor Networks[J]. ACTA Electronica Sinica, 2007,36(7):1291-1296.
41. Wang Xue, Wang Sheng, Ma Junjie. Dynamic Sensor Deployment Strategy Based on Virtual Force-Directed Particle Swarm Optimization in Wireless Sensor Networks[J]. ACTA Electronica Sinica, 2007,35(11):2038-2042.
42. Wang Yanli, An Shiquan. Research on the Coverage of Wireless Sensor Network[J]. Chinese Journal of Sensors and Actuator, 2005,18(2):307-312.
43. Zhou Tong, Hong Bingrong, Wang Yanqin. Fast Self-deployment Algorithm of Mobile Sensor Network[J]. Journal of Harbin Institute of Technology, 2005,37(7):870-872.