成绩：

西安建筑科技大学华清学院

毕业设计 (论文)英文翻译

院 （系）： 机械电子工程系

专业班级： 自动化0702

毕 业 设 计论 文 方 向 ： 单片机

翻 译 文 章 题 目 ：基于网络共享的无线传感网络设计

学生姓名： 程龙娜

学 号： 0706010237

指导教师： 赵敏华

2011年 4月 11日

基于网络共享的无线传感网络设计

摘要：无线传感器网络是近年来的一种新兴发展技术，它在环境监测、农业和公众健康等方面有着广泛的应用。在发展​​中国家，无线传感器网络技术是一种常用的技术模型。由于无线传感网络的在线监测和高效率的网络传送，使其具有很大的发展前景，然而无线传感网络的发展仍然面临着很大的挑战。其主要挑战包括传感器的可携性、快速性。我们首先讨论了传感器网络的可行性然后描述在解决各种技术性挑战时传感器应产生的便携性。我们还讨论了关于孟加拉国和加利

尼亚州基于无线传感网络的水质的开发和监测。

关键词：无线传感网络、在线监测

1.简介

无线传感器网络，是计算机设备和传感器之间的桥梁，在公共卫生、环境和农业等领域发挥着巨大的作用。一个单一的设备应该有一个处理器，一个无线电和多个传感器。当这些设备在一个领域部署时，传感装置测量这一领域的特殊环境。然后将监测到的数据通过无线电进行传输，再由计算机进行数据分析。这样，无线传感器网络可以对环境中各种变化进行详细的观察。无线传感器网络是能够测量各种现象如在水中的污染物含量，水灌溉流量。比如，最近发生的污染涌流进中国松花江，而松花江又是饮用​​水的主要来源。通过测定水流量和速度，通过传感器对江水进行实时监测，就能够确定污染桶的数量和流动方向。

不幸的是，人们只是在资源相对丰富这个条件下做文章，无线传感器网络的潜力在很大程度上仍未开发，费用对无线传感器网络是几个主要障碍之一，阻止了其更广阔的发展前景。许多无线传感器网络组件正在趋于便宜化（例如有关计算能力的组件），而传感器本身仍是最昂贵的。正如在在文献[5]中所指出的，成功的技术依赖于共享技术的原因是个人设备的大量花费。然而，大多数传感器网络研究是基于一个单一的拥有长期部署的用户，模式不利于分享。该技术管理的复杂性是另一个障碍。

大多数传感器的应用，有利于这样的共享模型。我们立足本声明认为传感器可能不需要在一个长时间单一位置的原因包括：（1）一些现象可能出现变化速度缓慢，因此小批量传感器可进行可移动部署，通过测量信号，充分捕捉物理现象（2）可能是过于密集，因此多余的传感器可被删除。（3）部署时间短。我们将会在第三节更详细的讨论。

上述所有假定的有关传感器都可以进行部署和再部署。然而有很多的无线传感器网络由于其实时监测和快速的网络功能可能被利用作为共享资源。其作为共同部署资源要求，需要一些高效的技术，包括对传感器的一些挑战，如便携性，流动频繁的传感器内的部署，这使我们在第四节将会有大的挑战。

在本文中，我们专注于作为共享的可行性设计的传感器网络。下面我们开始
阐述传感网络在孟加拉国和加利福尼亚州的水质检测中的应用。

2．无线传感网络在水质监测中的应用

无线传感器网络是通过把小型计算机设备连接到各式传感器和无线电而组成的。这些设备自适应的形成特殊网络（暂时的点对点网络），通过无线方式对所处环境进行监测、处理。其硬件和软件的设计非常低功耗以此达到长期在现场部署的目的，即此种部署在所处环境中人为干预性小。设备大小通常从四分之一个个人数据处理机到类似一个个人数据处理机的装置那么大。在一般情况下，资源可用性和功耗与设备大小是相一致的。例如，虽然资源可用性在很大程度上取决于传感器的功耗，但是低功率节点（通常称为微尘）用两节AA电池可以运行大约一二个月。

传感器网络提供密集的空间和时间上的采样。此种取样即使是在偏远和难以到达的地方均可采样。因此，它是对于在时间上和空间上要求精确采集最适用的网络技术。例如无线传感网络在土壤中的应用就是个很好的例子。因为土壤环境在空间上是多样性的，需要精确的时间上的采样。对于突然发生的变化都会被精确的采样及时记录下来。

事实上，无线传感器网络是一种低功耗的网络技术，对于一些发达地区其作为一种新兴技术适用性更为广泛。此外，对于公共健康方面的应用极为重要。例如，参考文献（17）阐述了人们对于水质的极高的关注度，“对水质的分析起初仍然是通过实验采样的办法将采集到的样本带回实验室进行研究。”这种类型的数据收集和分析通常是非常耗时的且大多是不准确的，并在许多情况下，错过了人们对于及时关注的焦点的分析。

我们参与了两项正在进行的关于地下水质监测的无线传感网络部署：一项系统是以了解孟加拉国地下水中砷的含量为主。另一项系统是通过研究孟加拉国地下水和土壤来监测硝酸盐的传播。

以上我们的部署都具有类似的设置。一个塔架，是由外围箱体式的无线设备组成，这些设备在土壤中通过长导线连接到嵌入式传感器。每个设备可以支持7个传感器，每个塔架都有多个设备。多路塔架被部署在目的地周围，以达到空间上垂直和水平的密集部署。这些设备将采集到的样本以无线方式传送给基站以供分析。该部署的基站是一个个人数据处理机类的设备，也可以是一种轻便电脑。它是通过由太阳能提供再充电的汽车电池来进行供电。为了能够获得外部数据，我们的基站使用Zigbee技术，或在Zigbee不可用时，使用GPRS网络。

在孟加拉国，在恒河三角洲的几千万人饮用了已被砷严重污染的地下水，如果被污染的水量一直持续，由砷引起的患病率和皮肤癌将大约每年分别增加两百万和一万例，由砷引起的癌症的死亡率每年将会大约增加三千例。

我们对于控制砷在地下水中的动态变化是难以完全了解的。在与孟加拉国的工程技术大学和麻省理工学院进行合作中，我们于2006年1月在靠近达卡的一个水稻地里部署了一个传感网络，目的是为了帮助确认这个假说成立。一个完整的塔架应该包含3部分完整的传感器（土壤湿度，温度，碳酸盐，钙，硝酸，氯，氧化还原电位，氨氮，pH值），每个部署都具有不同的深度（在地面以下1，1.5，2米），在此基础上的压力传感器用来监测水的深度。

在干旱地区和半干旱地区水的短缺和不断增加的对于水资源的消耗已经促进人们重新再利用被处理过的废水。尽管对于水资源的再利用人类收获了很多益处，但是已被处理的废水对于人类的健康和环境质量仍然存在着显而易见的危害。解决这些危害需要进行自动的分布式的观测和控制灌溉水量，查出它所传输的污染物，包括暂停处理的或是还未处理的污染物，胶状污染物，药物，有机碳，挥发性有机化合物，治病微生物，营养素例如氮或磷。在加利福尼亚的帕姆代尔，一个水质再利用现场是为测试土壤湿度，温度和硝酸盐的传感器网络而被用作的试车台。此网络集合了两个方面：第一，确保此环境正在被监测，第二，提供对水质控制的反馈，从而达到优化水流量和减少化学物质渗透到地下。这种现场也可以被用来在对孟加拉国进行部署前对软件，传感器和硬件的测试。

3．传感器共享技术

对于传感器网络数据收集，即使是最小的传感器资源，其共享也将让许多人受益。我们相信以下三种技术方法特别适用于传感器共享：（1）从一系列小型传感器大范围部署到精确仿真。（2）从密集部署到稀疏部署逐渐移动冗余传感器（3）在一些可能的地区缩短部署周期。在这里，我们更详细地描述这些场景，包括我们自己和别人在执行有关的或支持的算法时的工作的调查。

(1)精确仿真

人类功能的移动性就是通过手动来模拟一个使用较少传感器的密集部署区。人们可以移动一个领域的一小套传感器，对密集空间收集数据。该技术将是只适合于可持续发展应用中，所关注的现象变化非常缓慢。

(2)密集到稀疏部署

一些传感器网络应用需要一个密集映射的环境。一旦传感器密集部署和细节的现象揭示，我们可以看到它可以捕获足够的资料较少的传感器，从而释放传感器部署在其他地方。这里，我们描述适用的工作是正在进行中的传感器网络社区。

(3)部署周期短

有些应用程序只需要短时间部署，因而​​对传感器的共享是种理想选择。我们在孟加拉的部署是一个带有部署周期短的应用例子。我们要收集数据，以验证有关昼夜变化的假设，所以我们希望数天时间来对数据进行分析。

4．挑战

许多挑战性技术的出现，是为了能够快速部署和移动传感器，主要因为迄今为止的工作主要集中在静态的，长期运行的部署中。

我们已经有了趋于密集化的目标，降低高密度部署使之稀疏，使周期短的部署趋于平衡，我们发现以下三个挑战是最恰当的。算法必须是具有人机通信功能的，对于人为错误是可以解决的。对于系统故障必须迅速查明，并最大限度地通过正确的数据进行接收。最后，系统必须迅速做出部署。

5．结论

无线传感器网络可视为一种工具，其对于可持续发展来说具有很好的潜力。如果我们视这种发展的无线传感网络技术为共享资源的话，它就可以得到技术社区的帮助。为了使无线传感器网络作为一种共享资源得到落实，我们确定了三个有希望的技术方法：精确仿真，从密集部署到稀疏部署，实施短周期部署。我们讨论了我们的工作部署，这些部署已证明了这些技术，描述了我们的过去和现在需要做哪些工作去面对即将出现的重大挑战。

Designing Wireless Sensor Networks as a

Shared Resource

for Sustainable Development

Abstract ：Wireless sensor networks (WSNs) are a relatively new and rapidly developing technology; they have a wide range of applications including environmental monitoring, agriculture, and public health. Shared technology is a common usa ge model for technology adoption in developing countries. WSNs have great potential to be utilized as a shared resource due to their on-board processing and ad-hoc networking capabilities, however their deployment as a shared resource requires that the technical community first address several challenges. The main challenges include enabling sensor portability– the frequent movement of sensors within and between deployments, and rapidly deployable systems– systems that are quick and simple to deploy.We first discuss the feasibility of using sensor net-works as a shared resource, and then describe our research in addressing the various technical challenges that arise in enabling such senso rportability and rapid deployment. We also outline our experiences in developing and deploying water quality monitoring wireless sensor networks in Bangladesh and California.

Key words: WSNs、on-board processing

1 Introduction

Wireless Sensor Networks (WSNs), networks of wirelessly connected sensing and computational devices, hold tremendous promise for many areas of development including public health,the environment, and agriculture. A single device has a processor, a radio, and several sensors. When a network of these devices is deployed in a field, the sensing devices measure particular aspects of the environment. The devices then communicate those measurements by radio to one another and to more powerful computers for data analysis. In this way, WSNs can provide detailed observations of various phenomena that occur in the environment.

WSNs are capable of measuring diverse phenomena such as contaminant levels in water, pollutants in the air, and the flow of water for irrigation. As an example of a potential application, consider the recent incident of contamination spilling into the Songhua river in China, the main source of drinking water for many people1. Determining rate of flow and sometimes direction of the river requires coordination of multiple sampling points. Sensors periodically taking samples at multiple locations along the river could determine the rate, quantity, and direction of contaminant flow using the distributed sensing and processing of a wireless sensor network.

Unfortunately, the potential of wireless sensor net-works for sustainable development2 remains largely untapped while they are designed primarily for relatively resource-rich application contexts. The cost of WSNs is one of several major barriers that prevents them from being leveraged for sustainable development applications. Many components of WSNs are becoming cheaper (e.g. computing power), but the sensors themselves remain the most expensive component3. As stated in [5], successful technology-based international development projects rely on shared technology due to excessive cost of personal devices. However, most research on sensor networks is based on long-term deployments owned by a single user, a paradigm not conducive for sharing. The complexity of technology management is another barrier. We use Grameen telecom as a successful model 4 in which the management and maintenance of shared hardware is centralized. We envision a sensor network much in the same light.

Many sensor network applications are conducive to such a shared model. We base this statement on the observation that sensors may not be required in a single location for extended periods of time for reasons including: (1) a phenomenon of interest may have a slow rate of change, thus a small number of sensors can be moved within a deployment, emulating the density required to suciently capture the physical phenomena, (2) the initial deployment may have been too dense, thus redundant sensors can be removed, and (3) the duration of the deployment may be short. We discuss these scenarios in more detail in Section 3.

All of the deployment scenarios mentioned above rest on the assumption that sensors can be easily deployed and re-deployed. While WSNs have great potential to be utilized as a shared resource due to their on-board processing and ad-hoc networking capabilities, their deployment as a shared resource requires that the technical community first address several challenges, including enabling sensor portability – the frequent movement of sensors within and between deployments, and rapidly deployable systems – systems that are quick and simple to deploy. This leads us to our major challenges in Section 4.

Clearly, the primary issues related to successful technology adoption are the social, policy, and logistical questions to be answered in order to enable equitable access and the design of culturally appropriate technology. Our experience, though relevant, is limited to our technical expertise. These challenges and others should be formulated more explicitly with the necessary diverse input from communities, activists, governments and NGOs.

In this paper we focus on justifying the technical feasibility of designing sensor networks as a shared technology (Section 3) and describing the technical challenges that must be addressed to enable WSNs as a shared technology (Section 4). We begin by describing our applications in water quality monitoring in Bangladesh and California (Section 2).

2 WSNs For Water Quality

Wireless sensor networks are made up of small computational devices connected to various sensors and wireless radios. The devices automatically and adaptively form ad-hoc networks (temporary point-to-point networks) over wireless radios to make decisions based on measurements of their environment. The hardware and software are designed to be extremely low power in order to enable long-term in-situ deployments, i.e. undisturbed deployments that are left in the environment with minimal human intervention. Device sizes commonly range from that of a quarter to a PDA-like device. In general, resource availability and power consumption are commensurate with size. For example, while it largely depends on the power consumption of the sensors, the lower-power nodes (often called motes) can run for approximately one month on 2 AA batteries.

Sensor networks provide dense spatial and temporal sampling even in remote and hard to reach locations. Thus, they are best applied to applications that need dense sampling in space and/or time. Soi applications are a good example, because the soi environment is heterogeneous across space, requiring dense spatial sampling. Abrupt changes can then be captured with a high temporal sampling rate.

The fact that WSNs are low power and wireless makes them appealing as a technology for developing regions, but in addition the dense sampling is crucial for public health applications. For example, [17] states that while water quality concerns can be extremely critical, “analysis is still primarily conducted in a laborious manner by physical collection of a sample that is analyzed back in a laboratory.” This kind of data collection and analysis is time consuming and mostly undirected, and in many instances misses the toxin events of interest.

We are involved with two ongoing WSN deployments related to groundwater quality: a system to understand the prevalence of arsenic in Bangladesh groundwater, and a system to monitor nitrate propagation through soils and ground water in California.

Both of our deployments have a similar setup. A pylon [10] (Figure 2) consists of an enclosure housing the small wireless devices which connect to groups of sensors embedded at multiple depths in the soil through long wires. Each device can support 7 sensors and there are multiple devices per pylon. Multiple pylons are deployed around the field to attain vertical and horizontal spatial density. The devices wirelessly transmit samples back to a base-station for analysis (Figure 1). The base-station in these deployments was a PDA-class device. It could also be a laptop. It is powered by a car-battery recharged using solar panels. To make data externally accessible, our base-station is connected using Zigbee or where Zigbee is unavailable, using a GPRS (i.e. cellular) network.

In Bangladesh, tens of millions of people in the Ganges Delta drink ground water that is dangerously contaminated with arsenic. If consumption of contaminated water continues, the prevalence of arsenicosis and skin cancer will be approximately 2,000,000 and 100,000 cases per year, respectively, and the incidence of death from cancer induced by arsenic will be approximately 3,000 cases per year [18].

A full understanding of the factors controlling arsenic mobilization to ground water is lacking. In a joint collaboration with scientists at the Bangladesh University of Engineering and Technology and MIT, we deployed a sensor network in January of 2006 in a rice field near Dhaka,Bangladesh in order to aid in validating this hypothesis. A full pylon contains 3 complete suites of sensors (soil moisture, temperature, carbonate, calcium, nitrate, chloride, oxidation-reduction potential, ammonium, and pH), each deployed at a different depth (1, 1.5, and 2 meters below ground), and a pressure transducer at the base to monitor water depth.

Water scarcity in arid and semi-arid regions and increasing demand on water supplies has stimulated interest in the reuse of treated wastewater. Despite the many benefits to irrigating with reclaimed water, there remain both real and perceived risks to human health and environmental quality stemming from residuals in the treated wastewater. Proactively addressing these risks requires automating the distributed observation and control of the irrigation water and the trace pollutants that it conveys, including suspended or dissolved solids (TDS), colloidal solids, pharmaceuticals, organic carbon, volatile organic compounds, pathogenic microorganisms, and nutrients such as nitrogen or phosphorus. A water reuse site in Palmdale, California is being used as a testbed for a sensor network with soil moisture, temperature, and nitrate sensors. The network focuses on two things: first, ensuring that environmental regulations are being met, and second, providing feedback to a water control system in order to optimize water flow and minimize chemical penetration into the subsurface. This site is also used to test the software, sensors, and hardware before deploying in Bangladesh.

3 Sensor Sharing Techniques

Sensor sharing will allow many people to benefit from sensor network data collection, even with minimal sensor resources. We believe the following three technical approaches are particularly suited for enabling sensor sharing for sustainable development:(1) moving a smaller number of sensors around in a deployment to emulate density, (2) gradually removing redundant sensors from a deployment to go from dense to sparse deployments, and (3) leveraging shorter deployment cycles where possible. Here we describe each of these scenarios in greater detail, including a survey of our own and others’ work in implementing related or supporting algorithms.

（1）Emulating Density

Human-enabled mobility can be used to manually emulate the effect of a dense deployment using fewer sensors. People can move a small set of sensors around in a field in order to collect data for a dense spatial map of the field. This technique will be appropriate only for sustainable development applications in which the phenomenon of interest changes very slowly, on the order of days or longer.

（2） Dense to Sparse Deployments

Some sensor network applications require a dense mapping of the environment. Once sensors are densely deployed and details of the phenomenon are revealed, we may see it is possible to capture sufficient information with fewer sensors, freeing sensors for deployment elsewhere. Here we describe applicable work which is ongoing in the sensor network community.

（3） Short Deployment Cycles

Some applications only require short-duration deployments and therefore are ideal for sensor sharing. Our deployment in Bangladesh is an example of an application with a short deployment cycle. We wanted to collect data to validate a hypothesis about diurnal variations, and so we wanted several days of data for analysis.

4 Challenges

Numerous technical challenges arise in order to be able to quickly deploy and move sensors, primarily because the work to date has largely focused on static, long-running deployments.

Given that we have the goals to emulate density, reduce dense deployments to sparse ones, and leverage short deployments cycles, we find the following three challenges to be the most pertinent. Algorithms must be interactive and robust to human error. Faults in the system must be quickly identified to maximize the amount of good data received.Finally, systems must be made to be rapidly deployable.

5 Conclusion

Wireless sensor networks have the potential to be a useful tool for sustainable development. This can be facilitated by the technical community if we focus on issues with developing wireless sensor net-works as a shared technology. In order to implement WSNs as a shared resource, we identified three promising technical approaches: emulating density, moving from dense to sparse deployments, and implementing short deployment cycles. We discussed our work on deployments that have demonstrated these techniques and described our past and ongoing work to address the major challenges which arise.

References

[1] The General Assembly. 2005 World Outcome Document ref: Resolution adopted by the General Assembly 60/1. 2005 World Summit Outcome. United Nations, October 2005.

[2] M. Batalin, W. Kaiser, R. Pon, G. S. Sukhatme, G. Pottie, Y. Yu, J. Gordon, M. H. Rahimi, and D. Estrin. Task allocation for event- aware spatiotemporal sampling of environmental variables. In IEEE/RSJ Intl. Conf. on Intelligent Robots and Systems, August 2005.

[3] M. Batalin, M. H. Rahimi, Y. Yu, D. Liu, A. Kansal, G. S. Sukhatme, W. Kaiser, M. Hansen, G. Pottie, M. Srivastava, and D. Estrin. Call and response: Experiments in sampling the environment. In Procs. of ACM SenSys, pp. 25–38, Nov 2004.

[4] D. Bornstein. The Price of a Dream. The University of Chicago Press, 1997.

[5] E. Brewer, M. Demmer, B. Du, M. Ho, M. Kam, S. Nedevschi, J. Pal, R. Patra, S. Surana, and K. Fall. The case for technology in developing regions. In IEEE Computer, June 2005.

[6] A. Deshpande, C. Guestrin, S. Madden, and J. Hellerstein. Model-driven data acquisition in sensor networks. In Proceedings of VLDB, 2004.

[7] A. Dhariwal, B. Zhang, B. Stauffer, C. Oberg, G. S. Sukhatme, D. A. Caron, and A. A. Requicha. Networked aquatic microbial observing system. In IEEE Intl. Conf. on Robotics and Automation, May 2006.

[8] Thomas C. Harmon. Microsensors to the model forecasts:Multiscale embedded networked sensing of nutrients in the watershed. Eos Trans. American Geophysical Union, Fall Meet. Suppl.86(52):H23G–08 (Abstract No), 2005.

[9] Thomas C. Harmon, Richard F. Ambrose, Robert M. Gilbert, Jason C. Fisher, Michael Stealey, and William J. Kaiser. High resolution river hydraulic and water quality characterization using rapidly deployable networked infomechanical systems (nims rd). CENS Technical Report No. 60, 2006.

[10] E. Haux, T.C. Harmon, N.D. Busek, T. Schoellhammer, J. Saez, J. Kim, Y. Park, and D. Estrin. Embedded network sensing of moisture and nitrate propagation during irrigation with reclaimed wastewater. In Proceedings of the Annual International Meeting of the American Society of Agricultural Engineers, August 2004.

[11] A. Krause, C. Guestrin, A. Gupta, and J. Kleinberg.Near-optimal sensor placements:Maximizing information while minimizing communication cost. In Proceedings of Information Processing in Sensor Networks, 2006.

[12] United Nations. WCED, Our Common Future. Oxford University Press, 1987.

[13] M. H. Rahimi, M. Hansen, W. Kaiser, G. S. Sukhatme, and D. Estrin. Adaptive sampling for environmental field estimation using robotic sensors. In IEEE/RSJ Intl. Conf. on Intelligent Robots and Systems, pp. 747–753, August 2005.

[14] M. H. Rahimi, R. Pon, W. Kaiser, G. S. Sukhatme, D. Estrin, and M. Srivastava. Adaptive sampling for environmental robotics. In IEEE Intl. Conf. on Robotics and Automation, pp. 3537–3544, 2004.

[15] N. Ramanathan, K. Chang, R. Kapur, L. Girod, E. Kohler, and D. Estrin. Sympathy for the sensor network debugger. In Proc. of Sensys,2005.

[16] R. Willett, A. Martin, and R. Nowak. Backcasting: adaptive sampling for sensor net-works. In Proceedings of Information Processing in Sensor Networks, 2004.

[17] X. Yang, K.G. Ong, W.R. Dreschel, K. Zeng, C.S. Mungle, and C.A. Grimes. Design of a wireless sensor network for long-term, in-situ monitoring of an aqueous environment. Sensors,2:455–472, 2002.

[18] W. Yu, C. M. Harvey, and C. F. Harvey.Arsenic in the groundwater in bangladesh: A geostatistical and epidemiological framework for estimating health effects and evaluating remedies. Water Resources Research, 39(6),2003.

[19] B. Zhang, G. S. Sukhatme, and A. A. Requicha. Adaptive sampling for marine microorganism monitoring. In IEEE/RSJ Intl. Conf. on Intelligent Robots and Systems, pp. 1115 – 1122,2004.

自动化英文翻译