1、畢業(yè)設(shè)計(jì)外文資料翻譯學(xué) 院： 專業(yè)班級(jí)： 學(xué)生姓名： 學(xué) 號(hào)： 指導(dǎo)教師： 外文出處：(外文)JonathanW.Hui美 An Extended Internet Architecture for Low-Power Wireless Networks-Design and Implementation,2008 附 件：1.外文資料翻譯譯文； 2.外文原文 指導(dǎo)教師評(píng)語：該英文資料選擇合理，與畢業(yè)設(shè)計(jì)論文相關(guān)度較高。專業(yè)術(shù)語、詞匯翻譯的基本準(zhǔn)確，體現(xiàn)了基本的專業(yè)英語應(yīng)用水平。翻譯工作按照基本規(guī)定，翻譯材料能與原文能保持一致，能比較正確地表達(dá)出原文意思，細(xì)節(jié)處理基本得當(dāng)。翻譯字、詞數(shù)基本滿足

2、要求。翻譯材料語句通順，基本符合中文的表達(dá)習(xí)慣。翻譯工作能夠按照基本規(guī)定，能夠達(dá)到考核、提高學(xué)生英語應(yīng)用水平的目的，基本完成要求。簽名： 2015年10月14日1.外文翻譯譯文第二章背景Sensornet研究初見端倪大概是在十年前。此后，sensornet領(lǐng)域與低功耗無線網(wǎng)絡(luò)及其應(yīng)用的內(nèi)在要求被理解地更加成熟顯著。在同一時(shí)期，無論是IETF和IEEE標(biāo)準(zhǔn)組織已經(jīng)取得顯著的進(jìn)展。我們已經(jīng)看到了引進(jìn)IEEE 802.15.4的，專為低功耗，低成本，以及sensornet應(yīng)用的大規(guī)模需求而設(shè)計(jì)的。我們還看到引入IPv6和支持IP網(wǎng)絡(luò)的快速增長(zhǎng)所需的必要的協(xié)議和機(jī)制。在這兩個(gè)環(huán)節(jié)和IP網(wǎng)絡(luò)技術(shù)，與科研

3、相結(jié)合的發(fā)展進(jìn)步顯著，是什么使得它在今天是可行的并且真正融入到了sensornets互聯(lián)網(wǎng)架構(gòu)。在本章中，我們列舉的每一項(xiàng)都是本文的基礎(chǔ)也是技術(shù)進(jìn)步的背景。但是在這里我們先描述幾個(gè)積極使用sensornets的應(yīng)用。2.1一種新的計(jì)算類戈登貝爾觀察到了一類新的計(jì)算出現(xiàn)大約每十年和創(chuàng)造計(jì)算機(jī)課9貝爾定律。每一個(gè)新類的計(jì)算往往代表通過技術(shù)進(jìn)步，允許更低的成本和更小的計(jì)算設(shè)備啟用了一個(gè)顛覆性的技術(shù)。到目前為止，我們已經(jīng)看到通過大型機(jī)，小型機(jī)，個(gè)人電腦，手持設(shè)備的發(fā)展，以及嵌入式計(jì)算。 Sensornets代表下一節(jié)課計(jì)算。在低成本和低功率微控制器和收音機(jī)的突破，sensornets啟用需要大量嵌入在

4、物理環(huán)境中的節(jié)點(diǎn)的新的應(yīng)用。2.1.1典型應(yīng)用在過去的十年中，許多應(yīng)用程序已被提出，部署，并進(jìn)行評(píng)價(jià)。這方面的經(jīng)驗(yàn)使我們更好地了解應(yīng)用需求，這直接轉(zhuǎn)化為網(wǎng)絡(luò)挑戰(zhàn)。這些應(yīng)用要求從更傳統(tǒng)的IP網(wǎng)絡(luò)不同。在本節(jié)中，我們提出了一些sensornet應(yīng)用程序提供一個(gè)更好的了解所固有的網(wǎng)絡(luò)挑戰(zhàn)。環(huán)境監(jiān)測(cè)環(huán)境監(jiān)測(cè)應(yīng)用包括在農(nóng)業(yè)中使用，以提供有關(guān)土壤條件，更大的信息sensornets的;生物研究，以更好地了解鳥類難以捉摸的行為109，161或森林內(nèi)的小氣候提供更多的理解163，170;并執(zhí)行火山181的地質(zhì)研究。這些應(yīng)用程序通常需要的環(huán)境數(shù)據(jù)（例如，溫度，占用）中央服務(wù)器的慢周期報(bào)告 - 荷蘭國(guó)際集團(tuán)。異步

5、事件也可以由對(duì)感興趣的事件（即出現(xiàn)異常例如，振動(dòng)）單個(gè)傳感器的報(bào)道。從服務(wù)器到節(jié)點(diǎn)偶爾通信所需的配置或時(shí)間同步。這些應(yīng)用程序通常潛伏期傳遞的數(shù)據(jù)容錯(cuò)和偶爾的損失往往是可以接受的。結(jié)構(gòu)監(jiān)測(cè)和狀態(tài)維修結(jié)構(gòu)監(jiān)測(cè)和狀態(tài)檢修應(yīng)用包括高頻采樣，報(bào)告和處理振動(dòng)數(shù)據(jù)，以幫助預(yù)測(cè)工廠的機(jī)器，橋梁，或建筑物97，99，187的失敗。在振動(dòng)信號(hào)的高頻分量需傳感器節(jié)點(diǎn)以相對(duì)高的頻率進(jìn)行采樣和通信的更多信息到中央服務(wù)器比環(huán)境監(jiān)測(cè)應(yīng)用。像環(huán)境監(jiān)測(cè)，報(bào)告潛伏期一般是寬松的。但輸送可靠性經(jīng)常更高由于信號(hào)處理算法的要求。一些應(yīng)用中需要嚴(yán)格的時(shí)間同步，以跨越在相同的結(jié)構(gòu)放置在不同的意義上點(diǎn)的節(jié)點(diǎn)相關(guān)的時(shí)間序列數(shù)據(jù)。從服務(wù)器到節(jié)點(diǎn)

6、偶爾通信是必需的，用于配置和常常為可靠的傳輸協(xié)議。家庭和樓宇自動(dòng)化家庭和樓宇自動(dòng)化應(yīng)用設(shè)法消除參與控制建筑環(huán)境155接線。例子包括無線照明開關(guān)來控制照明燈具的單個(gè)或成組，無線恒溫控制空調(diào)系統(tǒng)，監(jiān)測(cè)門位置，入住，煙霧報(bào)警器，即使加上電源控制，以降低整體能源消耗的能源監(jiān)控設(shè)備的安全系統(tǒng)。像環(huán)境監(jiān)測(cè)，家庭和樓宇自動(dòng)化往往涉及到異步事件的環(huán)境數(shù)據(jù)的緩慢定期報(bào)告。然而，家庭和樓宇自動(dòng)化在幾個(gè)方面有所不同。首先，通信往往是在自然界雙向的，如自動(dòng)化應(yīng)用通常涉及既感測(cè)和致動(dòng)。第二，通信延遲必須發(fā)生在人類的時(shí)間尺度（例如，來自一個(gè)光開關(guān)的光的控制）。第三，節(jié)點(diǎn)可以與多個(gè)節(jié)點(diǎn)同時(shí)進(jìn)行通信，以支持用于控制多個(gè)致動(dòng)器

7、的單個(gè)輸入裝置（例如，光鎮(zhèn)流器在商業(yè)大廈的陣列）。最后，設(shè)備可以與其它鄰近設(shè)備直接通信，除了中央控制器。直接的溝通減少了等待時(shí)間和通信費(fèi)用，控制器和執(zhí)行器往往有空間位置（例如，光開關(guān)，在同一個(gè)房間的光）。工業(yè)自動(dòng)化工業(yè)自動(dòng)化應(yīng)用涉及監(jiān)測(cè)和工廠機(jī)器的控制，或用于制造項(xiàng)目的進(jìn)程。工業(yè)監(jiān)控應(yīng)用有不同程度的網(wǎng)絡(luò)要求。大多數(shù)應(yīng)用要求的傳感器值的緩慢定期報(bào)告即除了報(bào)警偶爾低延遲通信延遲容忍。有些應(yīng)用程序可能需要?dú)v史日志定期卸載。通常情況下，控制是通過在循環(huán)人類，其檢查需要由中央服務(wù)器和問題檢索的數(shù)據(jù)命令來個(gè)別的致動(dòng)器提供。閉環(huán)控制可以通過編程邏輯來實(shí)現(xiàn)進(jìn)入中央收集點(diǎn)。一些設(shè)想用于需要快速響應(yīng)時(shí)間的閉環(huán)控制

8、使用sensornets的，這可能涉及鄰近節(jié)點(diǎn)的直接通信，以滿足延遲要求。像所有到目前為止，我們已經(jīng)討論過的應(yīng)用程序，節(jié)點(diǎn)的物理位置是靜態(tài)的。然而，工業(yè)自動(dòng)化應(yīng)用通常需要傳遞的數(shù)據(jù)更高的可靠性和可能涉及延遲寬容和低延遲的通信。資產(chǎn)追蹤Sensornets已提出協(xié)助提供有關(guān)移動(dòng)資產(chǎn)的位置，連續(xù)的信息和有關(guān)檢測(cè)到該資產(chǎn)本身的信息。資產(chǎn)追蹤的概念涵蓋了多個(gè)行業(yè)，包括醫(yī)院和保健監(jiān)控各種醫(yī)療機(jī)械，航運(yùn)，物流的位置來定位包裝或者容器以及核查的高位置提供有關(guān)其內(nèi)容或安全應(yīng)用檢測(cè)的信息價(jià)值的物品。不像到目前為止，我們已經(jīng)討論了其他應(yīng)用程序，資產(chǎn)跟蹤涉及大量的移動(dòng)節(jié)點(diǎn)。移動(dòng)節(jié)點(diǎn)通?；谂c已公知的物理位置的固定節(jié)

9、點(diǎn)的無線電連接的推斷位置。這些固定的節(jié)點(diǎn)也形成用于移動(dòng)節(jié)點(diǎn)的通信基礎(chǔ)設(shè)施，并且通常具有與其他固定的節(jié)點(diǎn)恒定的網(wǎng)絡(luò)連接，但移動(dòng)節(jié)點(diǎn)僅可以根據(jù)它們是否是在基礎(chǔ)設(shè)施的范圍有間歇性連接。使用這些固定的節(jié)點(diǎn)，該移動(dòng)節(jié)點(diǎn)可以卸載的傳感器值的歷史記錄（例如，溫度，濕度和振動(dòng)）。城市網(wǎng)絡(luò)城市網(wǎng)絡(luò)是很像的資產(chǎn)跟蹤一組固定的基礎(chǔ)設(shè)施節(jié)點(diǎn)的部署和移動(dòng)節(jié)點(diǎn)通過基礎(chǔ)設(shè)施的通信時(shí)，它們的連接47，80。城市網(wǎng)絡(luò)可以承載許多應(yīng)用，其中包括故障檢測(cè)路燈或地下水管，監(jiān)測(cè)空氣質(zhì)量，協(xié)助在路邊停車場(chǎng)的應(yīng)用程序，并提供信息，并在流量的應(yīng)用提高效率。愿景是跨整個(gè)城市及其基礎(chǔ)設(shè)施部署城市網(wǎng)絡(luò)。2.1.2應(yīng)用要求從我們?cè)?.1.1節(jié)已經(jīng)確

10、定了應(yīng)用，也有脫穎而出sensornets從其他更傳統(tǒng)的網(wǎng)絡(luò)三個(gè)基本要求：（一）以物理空間嵌入，（二）大量節(jié)點(diǎn)，（三）總成本低的所有權(quán)。嵌入在物理空間這sensornets旨在解決的應(yīng)用包括接近感或動(dòng)作點(diǎn)節(jié)點(diǎn)的具體位置。不像在更傳統(tǒng)的網(wǎng)絡(luò)節(jié)點(diǎn)，節(jié)點(diǎn)sensornet往往不能放置在電源或網(wǎng)絡(luò)連接充足或方便。他們?cè)谖锢砜臻g嵌入放置在sensornet嚴(yán)格的要求。首先，必須sensornets無線操作，由自治區(qū)源供電，并使用無線電通信。無線操作給出嵌入在適當(dāng)?shù)奈恢胹ensornet節(jié)點(diǎn)和支持移動(dòng)應(yīng)用程序所需的靈活性。第二，物理位置往往限制sensornet節(jié)點(diǎn)的大小。自主電源（例如，電池）通常是系統(tǒng)

11、的最大組成部分，和尺寸限制直接轉(zhuǎn)換成有限的電源。物理約束可能限制電池的尺寸或通過太陽能電池板，振動(dòng)或其他方法的能力，以收集能量。節(jié)點(diǎn)大數(shù)sensornets的承諾是，他們能提供的傳感及驅(qū)動(dòng)點(diǎn)的高密度。單個(gè)應(yīng)用程序部署可以具有節(jié)點(diǎn)的數(shù)百或數(shù)千，并且通常根據(jù)單個(gè)管理域中操作。這與大多數(shù)其他無線網(wǎng)絡(luò)。如今，WiFi網(wǎng)絡(luò)是由筆記本電腦和手持設(shè)備的位置信息節(jié)點(diǎn)給用戶的比例仍然接近一對(duì)一為主。低功率無線網(wǎng)絡(luò)，如藍(lán)牙，定位到一個(gè)靠近七個(gè)比。然而，對(duì)于sensornets開發(fā)的IEEE 802.15.4標(biāo)準(zhǔn)靶向數(shù)千一個(gè)完全不同的操作點(diǎn)為一個(gè)?？傮w擁有成本較低隨著大量的節(jié)點(diǎn)，實(shí)施，部署和維護(hù)sensornet的

12、每個(gè)節(jié)點(diǎn)的成本必須是合理的。無線操作的要求是，必須盡量減少安裝成本加強(qiáng)。約束成本在每個(gè)節(jié)點(diǎn)（例如，能量容量和存儲(chǔ)器資源）限制的資源。為了限制維護(hù)成本，節(jié)點(diǎn)必須無人操作上適度的電源多年的壽命。還需要簡(jiǎn)單的配置和管理功能，以減少安裝和管理成本。2.2網(wǎng)絡(luò)挑戰(zhàn)他在物理空間，大量的節(jié)點(diǎn)，以及較低的總體擁有成本嵌入的三個(gè)基本應(yīng)用要求強(qiáng)加一大套網(wǎng)絡(luò)挑戰(zhàn)。低功率無線電臺(tái)有更短的范圍，更高的損失率，更動(dòng)態(tài)鏈路質(zhì)量比高功率無線電如WiFi。有限的內(nèi)存限制路由狀態(tài)，轉(zhuǎn)發(fā)緩存和鄰居信息的每個(gè)節(jié)點(diǎn)可以維持量。不可預(yù)知的部署屬性要求，具有廣泛的節(jié)點(diǎn)密度擴(kuò)展敏捷協(xié)議。在本節(jié)中，我們將討論sensornets固有的網(wǎng)絡(luò)挑戰(zhàn)

13、。在其核心，一個(gè)sensornet節(jié)點(diǎn)是由一個(gè)微控制器（MCU），低功率無線收發(fā)機(jī)，以及一個(gè)電源，如圖2.1。由于功耗和成本的限制，MCU和大多數(shù)sensornet應(yīng)用中使用的無線電設(shè)備相對(duì)簡(jiǎn)單，比在其他計(jì)算類提供較少的資源。通過實(shí)施減少功能，設(shè)備繪制工作和睡眠模式下的功率和更便宜的生產(chǎn)。簡(jiǎn)單性也允許MCU和無線電快速睡眠和活動(dòng)模式，實(shí)現(xiàn)整個(gè)系統(tǒng)的低功耗運(yùn)行的關(guān)鍵屬性之間進(jìn)行轉(zhuǎn)換。最sensornet應(yīng)用要求MCU和無線電都小于1的利用率來滿足其能量預(yù)算。MCURadioMCURadioPower Supply圖2.1：Sensornet節(jié)點(diǎn)架構(gòu)一個(gè)典型的sensornet節(jié)點(diǎn)是由一個(gè)微控制器

14、，低功率無線收發(fā)信機(jī)，和一個(gè)電源。實(shí)現(xiàn)最低的系統(tǒng)功耗，大部分硬件模塊必須保持休眠狀態(tài)，以滿足上盡可能。時(shí)，必須使用的硬件模塊，它應(yīng)該醒來，執(zhí)行所需的工作，并返回到盡可能快地睡覺。在下面的章節(jié)中，我們提出代表的MCU和sensornets用來提供一個(gè)更好地了解硬件限制收音機(jī)。我們還討論了從這種限制產(chǎn)生特定的網(wǎng)絡(luò)挑戰(zhàn)。2.外文原文Chapter 2BackgroundSensornet research began to take shape nearly a decade ago. Since then, the sensornet field has maturedsignificantlyw

15、ithbetterunderstandingsoftheinherentrequirementsoflow-powerwirelessnetworks and their applications. During that same time period, both the IETF and IEEE standards organizations have made significant advances. have seen the introduction of IEEE 802.15.4, specifically designed for the low low cost, an

16、d large scale demands of sensornet applications. have also seen the introduction of IPv6 and the necessary protocols and mechanisms needed to support the rapid growth of IP networks. Significant advances in both link and IP networking technologies, combined with research advances, are what makes it

17、feasible to truly integrate sensornets into the Internet architecture In this chapter, we providebackgroundoneachoftheadvancementsthatserveasafoundationforthisdissertation. Butfirst,we startwithdescribinganumberofapplicationsthatmotivatetheuseofsensornets.2.1 A New Class of ComputingGordonBellobserv

18、edthatanewclassofcomputingemergesabouteverydecadeandcoinedBells LawofComputerClassesHYPERLINK l _bookmark2339. Eachnewclassofcomputingoftenrepresentsadisruptivetechnologyenabled bytechnicaladvancesthatallowlowercostandsmallercomputingdevices. Sowehaveseenthe progression through mainframes, minicompu

19、ters, personal computers, handheld devices, and embeddedcomputing. Sensornets represent the next class of computing. With breakthroughs in low-cost and low-powermicrocon- trollers and radios, sensornets enable new applications that require large numbers of nodes embedded in the physicalenvironment.A

20、pplicationsOver the past decade, a number of applications have been proposed, deployed, and evaluated. This experience gives us a better understanding of the application requirements, which translate directly into networking challenges. These application requirements differ from more traditional IP

21、networks. In this section, we present a number of sensornet applications to provide a better understanding of the inherent networkingchallenges.Environmental MonitoringEnvironmental monitoring applications include the use of sensornets in agriculture to provide greaterinformationaboutsoilconditions;

22、biologicalstudiestoachieveabetterunderstandingofthebehavior of elusive birds HYPERLINK l _bookmark325109, HYPERLINK l _bookmark373161 or provide greater understanding of microclimates within forests HYPERLINK l _bookmark375163, HYPERLINK l _bookmark381170; and to perform geological studies of volcan

23、os HYPERLINK l _bookmark391181. These applications typically require slow periodic report- ing of environmental data (e.g., temperature, occupancy) to a central server. Asynchronous events may also be reported by individual sensors for interesting events (e.g., vibrations that appear abnormal). Occa

24、sional communication from the server to the nodes is required for configuration or time synchronization. These applicationsaretypicallylatencytolerantandoccasionallossindelivereddataisoftenacceptable.Structural Monitoring and Condition-Based MaintenanceStructural monitoring and condition-based maint

25、enance applications involve high frequency sam- pling, reporting, and processing of vibration data to help predict failures of factory machines, bridges, or buildings HYPERLINK l _bookmark31597, HYPERLINK l _bookmark31699, HYPERLINK l _bookmark397187. High frequency components in the vibration signa

26、l require sensor nodes to sample at relatively high frequencies and communicate more information to a central server than environmental monitoring applications. Like environmental monitoring, reporting latency is generally lenient. Butdelivery reliability is often higher due to the requirements of s

27、ignal processing algorithms. Some applicationsrequire tight time synchronization to correlate time series data across nodes placed at different sense points on the same structure. Occasional communication from the server to the nodes is required for configuration and often for reliable transportprot

28、ocols.Home and Building AutomationHome and building automation applications seek to eliminate the wiring involved in controlling the building environment HYPERLINK l _bookmark367155. Examples include wireless lighting switches to control individual or large groups of lighting fixtures, wireless ther

29、mostats to control systems, security systems that monitor door positions, occupancy, smoke alarms, and even energy monitoring devices coupled with power control to reduce overall energy consumption. Like environmental monitoring, home and building automationoften involves the slow periodic reporting

30、 of environmental data with asynchronous events. home and building automation differ in several ways. First, communication tends to be bidirectional in nature, as automation applications often both sensing and actuation. Second, communication latencies must occurathumantimescales(e.g.,controlofaligh

31、tfromalightswitch). Third,nodesmaycommunicatewith multiplenodessimultaneously,tosupportasingleinputdevicethatcontrolsmultipleactuators(e.g.,anarray of light ballasts in a commercial building). Finally, devices may communicate directly with other nearby devices, in addition to a central controller. D

32、irect communication reduces latency and communication costs, ascontrollersandactuatorstendtohavespatiallocality(e.g.,alightswitchandalightinthesameroom).Industrial AutomationIndustrial automation applications involve the monitoring and control of plant machinery or the processes used to manufacture

33、items. Industrial monitoring applications have varying degrees of network requirements.Most applications require slow periodic reporting of sensor values that is latency tolerant in addition to occasional low-latency communication of alarms. Some applications may require periodic offloading of histo

34、rical logs. Often, control is served by a human in the loop, which inspects data retrieved by a central server and issues commands to individual actuators as needed. Closed-loop control may be implemented by programming logic into the central collection point. Some envision the use of sensornets for

35、 closed-loop control that require quick response times, and this may the direct communication of nearby nodes to satisfy the latency requirements. Like all of the applications discussed so the physicalplacementofnodesarestatic. industrialautomationapplicationsgenerallyrequiregreater reliabilityforde

36、livereddataandmaybothlatency-tolerantandlow-latencycommunication.Asset TrackingSensornets have been proposed to assist in providing continuous information about the locationof mobile assets and sensed information about the asset itself. The concept of asset tracking spans a number of industries,incl

37、udinghospitalsandhealthcaretomonitorthelocationofvariousmedicalmachines,shipping andlogisticstolocatepackagesorcontainersandprovidesensedinformationabouttheircontents,orsecurity applicationsforverifyingthelocationofhighvalueitems. Unlikeotherapplicationswehavediscussedsoasset tracking involves a lar

38、ge number of mobile nodes. The mobile nodes typically infer location based on radioconnectivitywithstationarynodesthathavewell-knownphysicallocations. Thesestationarynodesalso form the communication infrastructure for mobile nodes and generally have constant network connectivity withotherstationaryn

39、odes,butmobilenodesmayonlyhaveintermittentconnectivitydependingonwhether or not theyre in range of the infrastructure. Using these stationary nodes, the mobile nodes may offload historical logs of sensor values (e.g. temperature, humidity, andvibration).Urban NetworksUrban networks are much like ass

40、et tracking in that a set of stationary infrastructure nodes are deployed and mobile nodes communicate through the infrastructure when they have connectivity HYPERLINK l _bookmark26847, HYPERLINK l _bookmark29980. Urbannetworkscanhostanumberofapplications,includingfailuredetectionforstreetlightsorun

41、derground water pipes, monitor air quality, assist in street parking applications, and provide information and increase efficiencyintrafficapplications. Thevisionisforurbannetworkstobedeployedacrossentiremunicipalities and theirinfrastructure.2.1.1 ApplicationRequirementsFromtheapplicationswehaveide

42、ntifiedinHYPERLINK l _bookmark9Section2.1.1,therearethreefundamentalrequirements that set sensornets apart from other more traditional networks: (i) embedding in physical space, (ii) large numbers of nodes, and (iii) low total cost ofownership.Embedding in Physical SpaceTheapplicationsthatsensornets

43、areintendedtoaddressspecificplacementofnodesnearthe sense or actuation point. Unlike nodes in more traditional networks, sensornet nodes often cannot be placed where power or network connectivity is plentiful or convenient. Their embedding in physical space places strictrequirementsonthesensornet. F

44、irst,sensornetsmustoperatewirelessly,beingpoweredbyautonomous sources and communicate using radios. Wireless operation gives the flexibility needed to embed sensornet nodesinappropriatelocationsandsupportmobileapplications. Second,thephysicallocationoftenlimitsthe size of sensornet nodes. The autono

45、mous power source (e.g., batteries) is often the largest component of the system, and size constraints translate directly to limited power sources. Physical constraints may limit the sizeofbatteriesortheabilitytoharvestenergythroughsolarpanels,vibration,orothermethods.Large Numbers of NodesThe promi

46、se of sensornets is that they can provide a high density of sense and actuation points. A single application deployment may have hundreds or thousands of nodes, and typically operate under a single administrative domain. This is in contrast to most other wireless networks. WiFi networks are dominate

47、d by laptops and handheld devices where the ratio of nodes to users remains near one to one.Lowerpower radionetworks,suchasBluetooth,targetarationearseventoone. theIEEE802.15.4 standarddevelopedforsensornetstargetedacompletelydifferentoperatingpointofthousandstoone.Low Total Cost of OwnershipWith la

48、rge numbers of nodes, the per-node cost of implementing, deploying, and maintaining a sensornet must be reasonable. The requirement for wireless operation is reinforced by the need to minimize installation costs. Constrained costs limit the resources at each node (e.g., energy capacity and memory re

49、sources). limit maintenance costs, nodes must operate unattended for multi-year lifetimes on modest power sources. Simple configuration and management capabilities are also needed to reduce installationand managementcosts.2.2Networking ChallengesThe three basic application requirements of embedding in physical space, large numbers ofnodes, and low total cost of ownership impose a large set of networking challenges. Low-power wireless radios have shorter range, higher loss rates, and more dynamic link qualities than high-power radios such as Wi