以安全、節能及遊憩為目的之車載網路系統---子計畫三：車載網路之位置感知服務---設計一個結合汽車及自行車行動導覽遊憩之系統( III )

行政院國家科學委員會專題研究計畫 期末報告

以安全、節能及遊憩為目的之車載網路系統--子計畫三：

車載網路之位置感知服務：設計一個結合汽車及自行車行

動導覽遊憩之系統(3/3)

計 畫 類 別 ： 整合型

計 畫 編 號 ： NSC 100-2219-E-009-002-

執 行 期 間 ： 100 年 08 月 01 日至 101 年 08 月 31 日

執 行 單 位 ： 國立交通大學資訊工程學系（所）

計 畫 主 持 人 ： 曾煜棋

共 同 主 持 人 ： 張馨文、吳宗修

報 告 附 件 ： 出席國際會議研究心得報告及發表論文

公 開 資 訊 ： 本計畫可公開查詢

中 華 民 國 101 年 11 月 03 日

中 文 摘 要 ： 隨著無線通訊與嵌入式微感知機電裝置(MEMS)技術的快速發

展，在行車環境利用無線通訊技術來發展車載感測網路(VSN)

已成為可能。一般車載感測網路是由數個被裝載在車輛上的

感測裝置，用來收集環境資訊或利用已偵測的資料來達成特

定目的，例如: 車輛追蹤、防碰撞系統、環境監測、行動監

控、以及車載安全等。因此，在第一年裡，本計畫開發一個

新型車載監視與感測系統，整合各式感測裝置技術、無線通

訊技術與車間通訊功能，以達成車載安全與車輛追蹤的目

的。在第二年的計畫執行中，我們提出了在道路上的每台車

輛均配備了影像攝影機並結合車牌辨識技術來識別可疑車輛

(例如贓車)，以及利用車載無線通訊介面與車間通訊方式來

協同追蹤可疑車輛和快速回報此發現給附近的警車。在第三

年計畫裡，我們提出了一個 VSN(vehicular sensor

network)的網路架構，利用移動性的車輛在微環境下來蒐集

空氣的品質(micro-climate monitoring) 。

中文關鍵詞： 汽車防盜、IEEE 802.15.4、監視系統、車載感測網路、車輛

追蹤、特定短距通訊、車牌辨識、車載監控網路、車載追

蹤、車載無線存取微環境偵測、機會式通訊、滲透式通訊、

車載感測網路、無線感測網路

英 文 摘 要 ： The rapid progress of embedded micro-sensing MEMS and

wireless communication technologies has made

vehicular sensor networks (VSNs) possible. A VSN

normally consists of a number of sensors placed on a

vehicle to collect environment data and utilizes

these sensed data for various purposes. Examples

include vehicle tracking, crash prevention, and

mobile surveillance. In the first year project,, we

are interested in taking advantage of the current

3G/3.5G mobile systems to enrich user interaction in

a VSN. Our goal is to develop a surveillance and

sensing system for car security and tracking

applications. In the second year project, we propose

that each vehicle employs a video camera to identify

suspicious vehicles (such as stolen cars) through

license plate recognition (LPR) technologies. In

addition, WAVE/DSRC-based radio interfaces are used

to cooperatively track the identified suspicious

vehicle and quickly report the discovery to nearby

police cars via vehicle-to-vehicle (V2V)

a VSN architecture to collect and measure air quality

for microclimate monitoring in city areas.

英文關鍵詞： Burglarproof, IEEE 802.15.4, Surveillance, Vehicular

Sensor Network, Vehicle Tracking, Dedicated Short

Range Communications, License Plate Recognition,

Vehicular Surveillance Network, Vehicle Tracking,

Wireless Access in Vehicular Environments

Micro-climate monitoring, Opportunistic communication,

Pervasive computing, Vehicular sensor network,

Wireless sensor network

行政院國家科學委員會補助專題研究計畫

■ 成 果 報 告

□期中進度報告

以安全、節能及遊憩為目的之車載網路系統－子計畫三：車

載網路之位置感知服務：設計一個結合汽車及自行車行動導

覽遊憩之系統(3/3)

計畫類別：□ 個別型計畫 ■ 整合型計畫

計畫編號：NSC 100－2219－E－009－002－

執行期間：100 年 8 月 1 日至 101 年 8 月 31 日

計畫主持人：曾煜棋

共同主持人：吳宗修、張馨文

計畫參與人員：陳怡秀、陳羿丞、吳建澄、邱俊瑋

成果報告類型(依經費核定清單規定繳交)：□精簡報告 ■完整報告

本成果報告包括以下應繳交之附件：

□赴國外出差或研習心得報告一份

□赴大陸地區出差或研習心得報告一份

■出席國際學術會議心得報告及發表之論文各一份

□國際合作研究計畫國外研究報告書一份

處理方式：除產學合作研究計畫、提升產業技術及人才培育研究計畫、

列管計畫及下列情形者外，得立即公開查詢

□涉及專利或其他智慧財產權，□一年□二年後可公開查詢

執行單位：國立交通大學資訊工程系

中 華 民 國 101 年 10 月 31 日

摘要

隨著無線通訊與嵌入式微感知機電裝置(MEMS)技術的快速發展，在行車環境利用無線通

訊技術來發展車載感測網路(Vehicular Sensor Network , VSN)已成為可能。一般車載感測網

路是由數個被裝載在車輛上的感測裝置，用來收集環境資訊或利用已偵測的資料來達成特

定目的，例如: 車輛追蹤、防碰撞系統、環境監測、行動監控、以及車載安全等。因此，

在第一年裡，本計畫開發一個新型車載監視與感測系統，整合各式感測裝置技術、無線通

訊技術與車間通訊功能，以達成車載安全與車輛追蹤的目的。在第二年的計畫執行中，我

們提出了在道路上的每台車輛均配備了影像攝影機並結合車牌辨識技術來識別可疑車輛

(例如贓車)，以及利用車載無線通訊介面與車間通訊方式來協同追蹤可疑車輛和快速回報

此發現給附近的警車。在第三年計畫裡，我們提出了一個VSN的網路架構，利用移動性的

車輛在微環境下來蒐集空氣的品質(micro-climate monitoring) 。

關鍵字：汽車防盜、

IEEE 802.15.4、監視系統、車載感測網路、車輛追蹤

、特定短

距通訊、車牌辨識、車載監控網路、車載追蹤、車載無線存取微環境偵測、機會式通訊、

滲透式通訊、車載感測網路、無線感測網路

Abstract

The rapid progress of embedded micro-sensing MEMS and wireless communication

technologies has made vehicular sensor networks (VSNs) possible. A VSN normally

consists of a number of sensors placed on a vehicle to collect environment data and

utilizes these sensed data for various purposes. Examples include vehicle tracking,

crash prevention, and mobile surveillance. In the first year project,, we are interested in

taking advantage of the current 3G/3.5G mobile systems to enrich user interaction in a

VSN. Our goal is to develop a surveillance and sensing system for car security and

tracking applications. In the second year project, we propose that each vehicle employs

a video camera to identify suspicious vehicles (such as stolen cars) through license

plate recognition (LPR) technologies. In addition, WAVE/DSRC-based radio interfaces

are used to cooperatively track the identified suspicious vehicle and quickly report the

discovery to nearby police cars via vehicle-to-vehicle (V2V) communications. In the

third year project, we propose

a VSN architecture to collect and measure air quality for

microclimate monitoring in city areas.

Keywords:

Burglarproof, IEEE 802.15.4, Surveillance, Vehicular Sensor Network,

Vehicle Tracking,

Dedicated Short Range Communications

, License Plate Recognition,

Vehicular Surveillance Network, Vehicle Tracking,

Wireless Access in Vehicular

Environments

Micro-climate monitoring, Opportunistic communication, Pervasive computing,

Vehicular sensor network, Wireless sensor network

目錄

一、前言………

7

二、研究目的

………...

8

三、研究方法

……….………

11

四、研究成果實驗

……….

18

五、結論與未來規劃

………..…

24

六、計畫成果自評

………..………

25

七、參考文獻

………..……..……….26

附錄：相關發表論文

1. L.-W. Chen, K.-Z. Syue, and Y.-C. Tseng, “VS

: A Vehicular Surveillance and Sensing

System for Security Applications”, The 6th IEEE International Conference on Mobile Ad Hoc

and Sensor Systems (MASS 2009), Oct. 12-15, 2009. (Receipt of Outstanding Demo Award)

2. S.-C. Hu, Y.-C. Wang, C.-Y. Huang, and Y.-C. Tseng, “A Vehicular Wireless Sensor

Network for CO

Monitoring”, The 8th Annual IEEE Conference on Sensors (Sensors 2009),

Oct. 25-28, 2009.

3. L.-W. Chen, K.-Z. Syue, and Y.-C. Tseng, “A Vehicular Surveillance and Sensing System

for Car Security and Tracking Applications”, The 9th ACM/IEEE International Conference on

Information Processing in Sensor Networks (IPSN 2010), Apr. 12-16, 2010.

4. L.-W. Chen, K.-Z. Syue, and Y.-C. Tseng, “An Implementation of a Vehicular Surveillance

and Sensing System for Car Security Applications”, The 3rd IEEE International Symposium

on Wireless Vehicular Communications Joint Telematics Workshop (in conjunction with VTC

2010-Spring), May 16-17, 2010.

5. L.-W. Chen, Y.-H. Peng, Y.-C. Tseng, and D.-C. Chang, “雙層式長鏈狀車載網路之高效

率資料收集與散佈機制”, The 6th Workshop on Wireless, Ad Hoc, and Sensor Networks

(WASN 2010), Sept. 2-3, 2010. (Recipient of Best Paper Award)

6. L.-W. Chen, Y.-H. Peng, and Y.-C. Tseng, “GoBike: A Group Communication System for

Bikers Based on Smart Phones”, Demonstration, ACM MobiCom, Sept. 20-24, 2010.

7. L.-W. Chen, Y.-H. Peng, and Y.-C. Tseng, “An Augmented Reality Based Group

Communication System for Bikers Using Smart Phones”, The 9th IEEE International

Conference on Pervasive Computing and Communications (PerCom 2011), Mar. 21-25, 2011.

8. L.-W. Chen, Y.-H. Peng, and Y.-C. Tseng, “An Infrastructure-Less Framework for

Preventing Rear-End Collisions by Vehicular Sensor Networks”, IEEE Communications

Letters, vol. 15, no. 3, pp. 358-360, Mar. 2011. (SCI, EI)

9. C.-C. Wu, L.-W. Chen, and Y.-C. Tseng, “Design and Implementation of a Bicycle Tour

Logging System Based on Smart Phones”, The Digital Content and Multimedia Applications

Conference, May 27, 2011.

10. L.-W. Chen, K.-Z. Syue, Y.-C. Tseng, and J.-H. Cheng, “Surveillance On-the-Road:

Suspicious Vehicle Tracking and Reporting Based on V2V Communications”, The 16th Mobile

Computing Workshop, June 17, 2011.

11. L.-W. Chen, P. Sharma, and Y.-C. Tseng, “Eco-Sign: A Load-Based Traffic Light Control

System for Environmental Protection with Vehicular Communications”, ACM International

Conference on Special Interest Group on Data Communication (SIGCOMM), Aug. 2011. (Demo

Session)

12. S.-C. Hu, Y.-C. Wang, C.-Y. Huang, and Y.-C. Tseng, “Measuring Air Quality in City Areas

by Vehicular Wireless Sensor Networks”, Journal of Systems and Software, Vol. 84, No. 11, pp.

2005-2012, Nov. 2011.

13. C.-C. Wu, L.-W. Chen, and Y.-C. Tseng, “Cooperative Localization for Power Saving in

Vehicular Long-thin Networks”, National Computer Symposium (NCS), Dec. 2011.

14. L.-W. Chen, J.-H. Cheng, Y.-C. Tseng, L.-C. Kuo, J.-C. Chiang, and W.-J. Lin, “LEGS: A

Load-balancing Emergency Guiding System Based on Wireless Sensor Networks”, IEEE

International Conference on Pervasive Computing and Communications (PerCom), Mar. 2012.

獲獎資料:

1. Recipient of Outstanding Demo Award in the 6th IEEE International Conference on Mobile

Ad-hoc and Sensor Systems, 2009. (IEEE MASS 2009 傑出展示獎)

2. Recipient of Best Paper Award in the 6th Workshop on Wireless, Ad Hoc, and Sensor

一、 前言

車載資通訊(Telematics)的蓬勃發展，對汽車業和資訊科技產業來說，是雙重的利多消息。

而車載資通訊研究範疇主要包含相連車輛之間或車輛與路邊裝置資訊的傳遞與運用的相關

技術，利用這些資訊建立智慧型的交通管理系統或是行車輔助系統，以提升行車安全及行

車效率為主要目的。隨著車載無線存取/特定短距通訊標準(Wireless Access in Vehicular

Environments /Dedicated Short Range Communications, WAVE/DSRC)和嵌入式監視系統技術

的快速發展，車輛可配備車上通訊裝置與影像攝影機來監控發生在道路上的各種事件，其

應用包括車輛安全、煞車示警和市區監控等。由於現在都會區的車輛數暴增，使得二氧化

碳的濃度急速增加，如何準確的監控二氧化碳濃度，相信是一個非常重要的課題。

二、 研究目的

第一年計畫:

我們設計了一個新型車載監視與感測系統，其可以達成車載安全與車輛追蹤之目的，

針對這樣的目的，我們把系統分成兩大模式，一為安全模式，另一為追蹤模式，對於安

全模式，我們利用目前的 3G/3.5G 行動通訊技術來監測與接收車輛上嵌入式系統的資

訊；對於追蹤模式，我們利用車牌辨識系統與 WAVE/DSRC 技術來持續追蹤可疑車輛，

並回報給相關部門。針對以上所提的兩種模式，以下是其個別分析簡述:

1. 安全模式：

在我們的設計規劃裡，當使用者停下車輛時，將會啟動此模式。對於

車內安全的應用，要發展一個監控與感測系統，一般傳統的車輛保護，通常會仰賴路邊的

攝影機來做影像記錄，對於這樣的設計，會存在兩個問題，第一，要識別目標車輛，需要

從大量的候選資料做篩選，這將造成很大的工作負荷。第二，因為目標車輛不是事先已知，

而且記錄的影像資訊又往往不夠清晰。更進一步，眾多的影像資訊，需要大量的人工勞力

來過濾與查詢。對此，我們提出一個以 3G/3.5G加強功能的VSN，稱為車載監視與感測系

統(Vehicular Surveillance and Sensing System, VS

)，在使用者端，僅需要一個 3G/3.5G行動

電話，在車輛端，要裝載CO

感測裝置、影像攝影機、3G/3.5G通訊模組和嵌入式系統開發

板。在嵌入式系統開發板上，其主要是請求命令與溝通協調各模組，VS

_{主要提供以下特色:}

(1) event-driven 模式，主要的事件來自於各模組偵測異常。

(2) 事件分為異常CO

空氣品質、車載防盜、異常聲音偵測。

(3) 支援文字或多媒體的互動。

僅當定義之異常事件發生時，影像攝影機將會被驅動來拍照或影像記錄，因此，當

沒有事件發生時，VS

_{可以避免不必要的影像記錄，而增進圖片/影像品質。對於應用場}

景包含了異常CO

偵測、車載防盜、異常聲音偵測，可用來驅動短簡訊服務(SMS)、多媒

體簡訊(MMS)或是互動式影像電話給車主。所以，VS

_{描述了一個新的車載安全與車載}

竊盜的模板。

2. 追蹤模式：

在我們的設計規劃裡，當使用者行駛期間，將會啟動這模式。對於追

蹤模式的應用，我們主要專注於追蹤贓車或可疑車輛並且通報給警車或是負責的車輛。針

對一般對於車載追蹤的做法，都要建置大量的 Roadside Unit (RSU)結合感測器於一般道路

旁，並與車載上的 On-Board Unit (OBU)通訊，但這樣將導致建置大量的 Roadside Unit

(RSU)，同時相對地提高了建置成本，但對於我們設計方法的最大好處是避免了建置 RSU

成本。此外，我們的方法最特別的地方是利用了 Wireless Access in Vehicular Environments

/Dedicated Short Range Communications (WAVE/DSRC)去增強車輛間在高速移動中的通訊能

力。

第二年計畫:

在本計畫中，我們對於可疑車輛追蹤與回報的問題定義如下，每個無線通訊介面的通訊

範圍為R，每一台車輛i持續識別其正前方的車輛v

是否為可疑車輛v

。我們的目標是設計出

高效率的機制來協同式追蹤已被識別的v

，並在追蹤期間於每個經過的交叉路口來回報v

最新的位置給v

，回報訊息m

是以多節點(Multi-hop)傳送方式導引至附近的v

，當v

接收到

m

後，便可重建v

的移動軌跡，以及儘快抵達v

所在位置以採取必要的處理措施，以下是我

們所設計之可疑車輛追蹤與回報機制的目標：

 追蹤工作換手(Tracking Handoff)：車道改變或路口轉向時可將追蹤工作換手至鄰近車輛

以持續追蹤v

。

 交叉路口偵測(Intersection Detection)：無需數位地圖輔助即可偵測出所經過的交叉路口

以回報v

最新位置給v

。

 位置回報廣播(Location Reporting)：根據鄰近車輛位置分佈來設計v

位置回報方式以減

少m

的重廣播數量。

 回報訊息導引(Message Guiding)：根據v

所經過之位置來將m

導引至最近的v

以減少位

置回報的訊息負擔(Overhead)。

第三年計畫:

能在市區隨時掌握二氧化碳的濃度和變動率，相信對於一些環保的研究課題是相當有幫

助的。由上述前言我們可以進而推出一個構想，就是利用車載網路來監控市區內二氧化碳

的濃度。在下面的報告裡，我們利用了一個微環境偵測的架構(micro-climate monitoring)， 也

就是說把要偵測的大範圍切成一些小區域，蒐集小範圍的二氧化碳，進而推估大範圍的二

氧化碳濃度。我們希望解決下面兩項問題：

1. 如何適當的調整車子的資料回報率(reporting rates)，以減少網路的流量。

2.

如何有效的利用動態式通訊(opportunistic communication)來減少通訊的負載。

我們提出了兩個演算法分別來解決上述的問題。

1.

為了要解決車子回報率的問題，首先，把要蒐集二氧化碳的區域(例如:新竹市)切成許

多的小方格(grid)，每塊小方格(grid)裡的車輛都要把二氧化碳濃度回報給伺服器。此種

演算法就是要推估每塊小方格裡面每台車輛的資料回報率(reporting rates)。

2. 由於車輛的移動具有隨機性，當一台車輛從一個小方格移動到另外一個小方格，如何

獲取當地的資料回報率？也因此我們提出了另外一個演算法來解決這個問題。

三、 研究方法

下面分別為三年計畫的研究方法，我們分別用條列式說明：

第一年計畫:(概述)

圖一顯示VS

_{系統架構，在車輛端，它包含了CO}

感測裝置、影像攝影機、WAVE/DSRC

通訊介面、3G/3.5G通訊模組和嵌入式系統開發板；在使用者端，則只需要一個 3G/3.5G行

動電話。為了說明VS

_{如何運作，我們將在以下用示意圖來描述車載安全、車載防盜、與車}

輛追蹤應用場景。

圖一. VS

_系統架構

第二年計畫:(概述)

圖二、可疑車輛追蹤模組與回報模組

如圖二所示，我們針對可疑車輛追蹤與回報問題提出一個由Tracking Module和Reporting

Module所構成的Infrastructure-less Framework，在Tracking Module方面，我們設計了Tracking

Handoff機制和Intersection Detection 機制來持續追蹤v

；在Reporting Module方面，我們設

計了Rebroadcast Decision機制、Intersection-guiding Search機制、以及Memory-based Backoff

第三年計畫:

圖三是針對演算法所提出的網路架構，此結構有下列元素構成。

圖三. 針對微環境偵測所提出的 VSN(Vehicular sensor network)網路架構

 Vehicular sensor：配備有二氧化碳偵測裝置的車輛，其中又分為External unit 和Central

unit。

1. External unit：External unit是放在車外的sensor裝置，主要是用來收集二氧化碳的

資料。

2.

Central unit：Central unit是放在車內的裝置，由cellular interface(2G/3G/3.5G) 、GPS

receiver以及兩個wireless interface構成。

 Other vehicle：沒有配備二氧化碳偵測裝置的車輛。

 GSM base station：2G網路基地台。

 GSM network：2G網路連線。

 Monitoring server：二氧化碳監控中心。

為了審查者能更詳細的聊解整個網路架構運作過程，我們用下列的情境模擬說明：

 Step1：假設有一台車輛(Vehicular sensor)已經收集到二氧化碳的資料。

 Step2：車輛外部的二氧化碳感應器會把資料傳給內部的Central unit。

 Step3：Central unit會根據通訊協定把資料傳給GSM base station。

 Step4：GSM base station再把資料傳給Monitoring server。

形成無線隨意網路(Ad hoc Network) ，這樣可以增加機會式通訊的效能。然而，每台車可

以藉由可以蒐集鄰近車輛的資訊，再把資訊傳送給二氧化碳監控中心。

在下 一 個 部 份的 核心 演算 法， 分成 DRR(Dynamic reporting rates) 和TOR(time-constraint

opportunistic relay)兩部分來介紹。下面先介紹此演算法的動機。

 DRR(Dynamic reporting rates)：如果要預測一個都市的二氧化碳濃度，除了考慮二氧

化碳的分布之外，還必須考慮各區域的車子密度，因為車輛會隨著時間而有所變動。

DRR演算法就是要在每個小方格(grid)裡去預測此區域的資料回報率(Reporting rates) 。

 TOR(time-constraint opportunistic relay) ：當一台車從原本的方格G1移動到另外一個

方格G2時，這台車可以幫忙G2裡面的車輛傳送資料，此時就可以降低G2方格裡的資料

回報率。

由於二氧化碳的分布和變動是動態的，如果要精確的預估二氧化碳濃度，除了根據過去

的數據資料來做參考還必須考慮下列兩點。

1.

當地區域的二氧化碳濃度。

2.

當地區域的車輛數。

下面部分開始介紹核心演算法。

 DRR(Dynamic reporting rates)：

DRR演算法主是要用來預估二氧化碳資料的回報率。首先，如果要預測一個大區域的

二氧化碳濃度，先把此大區域切成許多小的區域(圖四) 。為了要準確的預測資料回報

率，DRR又可再細分為兩種。

 Variation-based scheme：以模擬情境圖(圖四)的範例來看，座標(2,4)和座標(2,5)

二氧化碳濃度偏高，因此，此地區的資料回報率會偏高；座標(1,5)和座標(2,6)的

二氧化碳濃度偏低，因此，此地區的資料回報率會偏低。所以我們提出了一個數

學式子，來推估該區域的資料回報量。

圖四. 情境模擬圖

這個數學式是一個線性的式子，利用一個變數和兩個常數來推估資料回報量。

r

= S

/ Vi

S

= a

×ƃ

+b

ƃ

：

資料量的變數

r

_：

每台車需要的回報量

S

：

該區域資料回報總量

a

_和b

：

過去經驗的常數

Vi

：

該區域的車輛總數

 Gradient-based scheme：上述的數學式子並沒有考慮到單位距離二氧化碳濃度的

變化量。假設有兩台車。他們之間距離很近，但是他們所在位置的二氧化碳濃度

差異很大，所以這兩台車的之間單位濃度變化量很大，這是一個非常重要的資訊。

因此我們提出另外一個利用梯度(gradient)的公式。

首先，算出任意X和Y兩點的斜率(單位距離濃度的變化) ，公式如下：

α(X,Y )=X-Y/dist(X,Y)

再來，推算出斜率總和，

R

和

R

代表收集到高濃度和低濃度的車輛數：

α

i

=Σ

∈

R

∈

R

α(X,Y) / R

×R

接下來，再用Variation-based scheme的公式重新計算每輛車的回報量：

r

= S

/ Vi

S

= a

_×α

+b

 TOR(time-constraint opportunistic relay) ：當一輛車子原本的小方格G1移動到另外一

個小方格G2時，可以藉由機會式通訊(opportunistic communication)來幫助G2內的車輛

傳送資料。也因此，我們定義了下面的演算法。

1.

首先，當車輛X進入新的方格G2時，會定期廣播一個封包(HELLO Packet) 。這個

封包括目前的方格ID還有一個

att

(attraction value)

數值。

att

=rnd

×wgt

rnd

：

0~1的隨機亂數

2.

當X進入新的方格G2時，X可以藉由External unit和 Central unit來和G2內部的車輛

Y互相交換訊息。

3.

當X和Y相遇時，而且

wgt

wgt

，X會把自己的資料量回報給Y，在傳送完畢後，

X會把

wgt

改回零，而Y會把原本的

wgt

再加上

wgt

。

4.

當X如果因為一些障礙(如:大樓)而失去衛星導航信號，X會藉由通訊裝置和相鄰節

點溝通來獲取自己的座標位置。

5.

最後，權重值不為零的點來傳送資料。

四、 研究成果和實驗

在第一年的研究成果裡，我們實做了VS

_{裝置間的模組區塊橋接介面(圖五)。微處理器在}

VS

裝置是採用ARM9 開發板(Mini2440)，配有 3.5” TFT LCD、一個 400MHz 32-bit RISC處

理器(ARM920T)、64MB SDRAM、64MB Nand Flash、2MB Nor Flash用於BIOS、三組序列

傳輸阜、10/100M Ethernet RJ-45，Mini2440(圖六)可以執行embedded Linux和WinCE發展各

式應用。

圖五. VS

模組區塊橋接介面

在第二年研究成果裡，們使用QualNet 5.0 網路模擬器並加入必要之修改來作模擬實驗

的效能評估，如圖七所示，實驗環境拓樸為 5 km

_{的城市區域，每個街區的大小為 1 km}

_。

所有車輛都均勻地散布在各個街道上，隨機選擇出一可疑車輛與一警車，每台車輛在路口

隨機選擇前進、右轉、左轉三個方向其中之一。表一為我們所使用的模擬實驗參數，並設

定t

= 1 s、t

= 10s、θ = 60

、τ= 1、ρ= 3、T = 30 s。圖七顯示在道路上不同的總車輛數

所造成之封包碰撞率。圖八顯示平均回報延遲時間比較。由此可知，我們的演算法效能較

好。

表一.模擬實驗參數

圖七.封包碰撞率比較

圖八. 平均回報延遲時間比較

在第三年研究成果裡，我們實做了一個偵測二氧化碳的硬體裝置(圖九)，圖的左方是監測

二氧化碳的板子，圖的右方是External unit和Central unit的裝置。圖十則是這篇論文的封包

傳輸格式。在模擬實驗方面，我們用C++和Matlab來驗證結果。表二是模擬參數，在模擬實

驗裡，假設每傳出去一個資料花費一塊錢。由圖十一我們可以發現Variation-based scheme

平均的花費會比Gradient-based scheme高，這是因為Gradient-based scheme有考慮到單位長度

的濃度變化，所以所花費會比較低一點。而在預估誤差方面，兩者則是差不多。圖十二我

們可以發現Variation-based scheme平均的花費會比Gradient-based scheme高，這是因為

Gradient-based scheme有考慮到單位長度的濃度變化，所以所花費會比較低一點。而在預估

誤差方面，兩者則是差不多。

圖九. 硬體實做裝置

圖十. 封包傳輸格式

五、 結論與未來規劃

在第一年的計畫成果中，我們已開發了新型車載監視與感測系統，並以其為基

礎發展了車載

安全應用、車載防盜應用、以及

車輛

追蹤應用。在第二年度的計畫執行

中，我們藉由車間通訊技術設計出一個無需搭配基礎建設的車載網路追蹤機制，其中包

含追蹤換手(Tracking Handoff Scheme)、路口偵測(Intersection Detection Scheme)、決定重

廣播設計(Rebroadcast Decision Scheme)、路口導引搜尋設計(Intersection-Guiding Search

Scheme)和記憶式 backoff 設計(Memory-Based Backoff Scheme)。在這些機制的運作下，

可以充分地減少網路上控制封包的負荷量和不必要的重廣播封包量。在第三年的計畫

裡，我們把二氧化碳的硬體裝置實做出來，並且利用車輛的移動性和隨機性來蒐集二氧

化碳濃度。先針對小區域的環境來回報，進而推估整個大區域的濃度。我們並且針對 VSN

提出了一個新的網路架構來做傳輸。為了考慮網路的傳輸量和車輛的移動性，我們也提

出了新的數學式子來做效能評比和做改善。最後，透過模擬來驗證我們的演算法。在三

年計畫執行期間，我們得到以下的成果：國內外會議與國際期刊論文發表 14 篇以及學術

獎項 2 件，未來我們將朝向實務系統和標準開發規劃。

六、計畫成果自評

在三年計畫的執行過程中，已完成下列數項成果：(1)車載環境CO

感測系統之設計與實

作，本成果己發表於IEEE Sensors 2009 國際研討會；(2)新型車載監視感測系統之設計與實

作，本成果己發表於IEEE MASS 2009 國際研討會，並獲大會頒發Outstanding Demo Award；

(3)在所實作之車載監視感測系統上開發出新型車輛安全應用，本成果己發表於VTC/WiVEC

Joint Telematics Workshop 2010 國際研討會；(4)在所實作之車載監視感測系統上開發出新型

車輛追蹤應用，本成果己發表於ACM/IEEE IPSN 2010 國際研討會；(5)雙層式長鏈狀車載

網路之高效率資料收集與散佈機制設計，本成果己發表於WASN 2010 研討會，並獲大會頒

發Best Paper Award；(6)GoBike自行車隊通訊系統之設計與實作，本成果己於ACM MobiCom

2010 國際研討會進行雛型系統展示；(7)在所實作之自行車隊通訊系統上開發出新型擴增實

境應用，本成果己發表於IEEE PerCom 2011 國際研討會；(8)車載感測網路中無需基礎設施

輔助之車輛防追撞機制設計，本成果己發表於IEEE Communications Letters國際期刊；(9)

以智慧型手機為基礎之單車旅遊紀錄系統設計與實作，本成果己發表於Digital Content and

Multimedia Applications Conference 2011 研討會；(10)以車間通訊為基礎之可疑車輛追蹤與

回報系統設計與實作，本成果己發表於Mobile Computing Workshop 2011 研討會；(11)基於

車載網路以環境保護為目的之交通號誌控制機制，本成果發表於ACM SIGCOMM 2011 國

際研討會；(12)以車載感測網路為基礎之都市空氣監控機制，本成果發表於Journal of

Systems and Software國際期刊；(13)在車載長鏈狀網路以節能為目的之協同式定位機制，本

成果發表於NCS 2011 研討會；(14)以無線感測網路為基礎之負載平衡緊急導引系統，本成

果發表於IEEE PerCom 2012 國際研討會。

七、參考文獻

[1] Allred, J., Hasan, A.B., Panichsakul, S., Pisano, W., Gray, P., Huang, J., Han, R., Lawrence,

D., Mohseni, K., 2007. Sensorflock: an airborne wireless sensor network of microair

vehicles. In: Proc. ACM International Conference on Embedded Networked Sensor

Systems , pp. 117–129.

[2] Bai, F., Sadagopan, N., Helmy, A., 2003. Important: a framework to systematically analyze

the impact of mobility on performance of routing protocols for adhoc networks. In:

INFOCOM 2003. Twenty-Second Annual Joint Conference of the IEEE Computer and

Communications. IEEE Societies, vol. 2 , pp. 825–835.

[3] Cao, G., Kesidis, G., Porta, T.L., Yao, B., Phoha, S., 2006. Purposeful mobility in tactical

sensor networks. Sensor Network Operations.

[4] Chebrolu, K., Raman, B., Mishra, N., Valiveti, P.K., Kumar, R., 2007. Luster: wireless

sensor network for environmental research. In: Proc. ACM International Conference on

Embedded Networked Sensor Systems , pp. 103–116.

[5] Chebrolu, K., Raman, B., Mishra, N., Valiveti, P.K., Kumar, R., 2008. Brimon: a sensor

network system for railway bridge monitoring. In: Proc. ACM International Conference on

Mobile systems, Applications, and Services , pp. 2–14.

[6] Eisenman, S.B., Miluzzo, E., Lane, N.D., Peterson, R.A., Ahn, G.S., Campbell, A.T., 2009.

Bikenet: a mobile sensing system for cyclist experience mapping. ACM Transactions on

Sensor Networks 6, 6:1–6:39.

[7] Google Maps, 2010. http://maps.google.com/.

[8] Gopakumar, A., Jocob, L., January 2008. Localization in wireless sensor networks using

particle swarm optimization. In: IET International Conference on Wireless, Mobile and

Multimedia Networks , pp. 227–230.

[9] H-550EV module, 2008. http://www.elti.co.kr/.

[10] He, T., Krishnamurthy, S., Stankovic, J.A., Abdelzaher, T., Luo, L., Stoleru, R., Yan, T., Gu,

L., Zhou, G., Hui, J., Krogh, B., 2006. VigilNet: an integrated sensor network system for

energy-efficient surveillance. ACM Transactions on Sensor Networks 2 (1), 1–38.

[11] Hull, B., Bychkovsky, V., Zhang, Y., Chen, K., Goraczko, M., Miu, A., Shih, E.,

Balakrishnan, H., Madden, S., 2006. Cartel: a distributed mobile sensor computing system.

In: Proc. ACM International Conference on Embedded Networked Sensor Systems , pp.

125–138.

[12] IEEE standard for information technology–telecommunications and information exchange

between systems–local and metropolitan area networks specific requirements part 15.4:

wireless medium access control (MAC) and physical layer (PHY) specifications for low-rate

wireless personal area networks (LRWPANs), 2006.

[13] Jennic JN5139, 2008. http://www.jennic.com/.

[14] Juang, P., Oki, H., Wang, Y., Martonosi, M., Peh, L., Rubenstein, D., 2002. Energyefficient

computing for wildlife tracking: design tradeoffs and early experiences with zebranet. ACM

SIGOPS Operating Systems Review 36 (5), 96–107.

[15] Kargupta, H., Bhargava, R., Liu, K., Powers, M., Blair, P., Bushra, S., Dull, J., Sarkar, K.,

Klein, M., Vasa, M., Handy, D., 2004. VEDAS: a mobile and distributed data stream mining

system for real-time vehicle monitoring. In: Proc. SIAM International Conference on Data

Mining , pp. 300–311.

[16] Lee, U., Zhou, B., Gerla, M., Magistretti, E., Bellavista, P., Corradi, A., 2006. Mobeyes:

smart mobs for urban monitoring with a vehicular sensor network. IEEE Wireless

Communications 13 (5), 52–57.

[17] Li, M., Liu, Y., 2007. Underground structure monitoring with wireless sensor networks. In:

Proc. International Symposium on Information Processing in Sensor Networks , pp. 69–78.

[18] Liu, K., Li, M., Liu, Y., Li, M., Guo, Z., Hong, F., 2008. Passive diagnosis for wireless

sensornetworks. In: Proc. ACM International Conference on Embedded Networked Sensor

Systems , pp. 113–126.

[19] Sheu, J.P., Chen, P.C., Hsu, C.S., 2008. A distributed localization scheme for wireless

sensor networks with improved grid-scan and vector-based refinement. IEEE Transactions

on Mobile Computing 7 (September (9)), 1110–1123.

[20] SIM300 module, 2008. http://www.sim.com/.

[21] Tseng, Y.C., Wang, Y.C., Cheng, K.Y., Hsieh, Y.Y., 2007. iMouse: an integrated mobile

surveillance and wireless sensor system. IEEE Computer 40 (6), 60–66.uPatch300 module,

2008. http://www.fastraxgps.com/.

monitoring. In: EURASIP Journal on Applied Signal Processing, vol. 2003, pp. 392–401.

[23] Wang, Y.C., Wu, F.J., Tseng, Y.C., 2010. Mobility management algorithms and applications

for mobile sensor networks. Wireless Commununications and Mobile Computing.

[24] Werner-Allen, G., Johnson, J., Ruiz, M., Lees, J., Welsh, M., 2005. Monitoring volcanic

eruptions with a wireless sensor network. In: Proc. European Workshop on Wireless Sensor

Networks , pp. 108–120. Werner-Allen, G., Lorincz, K., Ruiz, M., Marcillo, O., Johnson, J.,

Lees, J., Welsh, M., 2006. Deploying a wireless sensor network on an active volcano. IEEE

Internet Computing 10, 18–25.

[25] Xu, N., Rangwala, S., Chintalapudi, K.K., Ganesan, D., Broad, A., Govindan, R., Estrin,

D., 2004. A wireless sensor network for structural monitoring. In: Proc. ACM.

[26] ARM, ARM920T. http://www.arm.com/products/CPUs/ARM920T.html.

[27] C. Sharp, S. Schaffert, A. Woo, N. Sastry, C. Karlof, S. Sastry, and D. Culler. Design and

implementation of a sensor network system for vehicle tracking and autonomous

interception. In Proceeedings of the Second European Workshop on Wireless Sensor

Networks, pages 93–107, 2005.

[28] D. Djenouri. Preventing vehicle crashes through a wireless vehicular sensor network. In 24

Biennial Symposium on Communications, pages 320–323, June 2008.

[29] FriendlyARM, Mini2440. http://www.friendlyarm.net.

[30] H-550EV CO

Sensor Module.

http://www.co2sensor.co.kr/new/eng/ndirco2-sensor-module-h550ev.htm.

[31] Jennic, JN5139. http://www.jennic.com.

[32] Y.-C. Tseng, Y.-C. Wang, K.-Y. Cheng, and Y.-Y. Hsieh. iMouse: An integrated mobile

surveillance and wireless sensor system. IEEE Computer, 40(6):60–66, 2007.

[33]

IEEE std. 802.11p/D4.0, Draft Amendment for Wireless Access in Vehicular

Environments (WAVE). Mar. 2008.

[34] K. G. Aravind, T. Chakravarty, M. G. Chandra, and P. Balamuralidhar. On the architecture

of vehicle tracking system using wireless sensor devices. In International Conference on

Ultra Modern Telecommunications & Workshops (ICUMT 2009), pages 1–5, Oct. 2009.

[35] M. Caceres, F. Sottile, and M.A. Spirito. WLAN-Based Real Time Vehicle Locating System.

In The 69

IEEE Vehicular Technology Conference (IEEE VTC-Spring 2009), pages 1–5,

Apr. 2009.

[36] M. Li, H. Zhu, Y. Zhu, and L. M. Ni. ANTS: Efficient Vehicle Locating Based on Ant

Search in ShanghaiGrid. IEEE Transactions on Vehicular Technology, 58(8):4088–4097,

Oct. 2009.

[37] M. Li, M.-Y. Wu, Y. Li. ShanghaiGrid: An Information Service Grid. Concurrency and

Computation: Practice and Experience, 18:111–135, 2006.

[38] P. Thammakaroon and P. Tangamchit. Adaptive Brake Warning System for Automobiles. In

8

International Conference on ITS Telecommunications (ITST 2008), pages 204–208, Oct.

2008.

[39] R. Lu, X. Lin, H. Zhu, and X. Shen. SPARK: A New VANET-Based Smart Parking Scheme

for Large Parking Lots. In The 28

IEEE Conference on Computer Communications (IEEE

INFOCOM 2009), pages 1413–1421, Apr. 2009.

[40] U. Lee, E. Magistretti, M. Gerla, P. Bellavista, and A. Corradi. Dissemination and

harvesting of urban data using vehicular sensing platforms. IEEE Transactions on Vehicular

Technology, 58(2):882–901, Feb. 2009.

[41] L.-W. Chen, K.-Z. Syue, and Y.-C. Tseng, “VS

: A Vehicular Surveillance and Sensing

System for Security Applications,” in IEEE International Conference on Mobile Ad-hoc and

Sensor Systems (IEEE MASS 2009), Oct. 2009.

[42] P. Thammakaroon and P. Tangamchit, “Adaptive Brake Warning System for Auto- mobiles,”

in The 8th International Conference on ITS Telecommunications (ITST 2008), pp. 204–208,

Oct. 2008.

[43] U. Lee, E. Magistretti, M. Gerla, P. Bellavista, and A. Corradi, “Dissemination and

harvesting of urban data using vehicular sensing platforms,” IEEE Transactions on

Vehicular Technology, vol. 58, no. 2, pp. 882–901, Feb. 2009.

[44] K. G. Aravind, T. Chakravarty, M. G.Chandra, and P. Balamuralidhar, “On the ar- chitecture

of vehicle tracking system using wireless sensor devices,” in International Conference on

Ultra Modern Telecommunications & Workshops (ICUMT), pp. 1–5, Oct. 2009.

[45] M. Caceres, F. Sottile, and M.A. Spirito, “WLAN-Based Real Time Vehicle

LocatingSystem,” in The 69th IEEE Vehicular Technology Conference (IEEE VTC-Spring

2009), pp. 1–5, Apr. 2009.

[46] R. Lu, X. Lin, H. Zhu, and X. Shen, “SPARK: A New VANET-Based Smart Park- ing

Scheme for Large Parking Lots,” in The 28th IEEE Conference on Computer

Communications (IEEE INFOCOM 2009), pp. 1413–1421, Apr. 2009.

[47] M. Li, H. Zhu, Y. Zhu, and L. M. Ni, “ANTS: Efficient Vehicle Locating Based on Ant

Search in ShanghaiGrid,” IEEE Transactions on Vehicular Technology, vol. 58, no. 8, pp.

4088–4097, Oct. 2009.

[48] M. Li, M.-Y. Wu, Y. Li, “ShanghaiGrid: An Information Service Grid,” Concurrency and

Computation: Practice and Experience, vol. 18, pp. 111–135, 2006.

[49] IEEE std, “802.11p/D4.0, Draft Amendment for Wireless Access in Vehicular Environments

(WAVE),” Mar. 2008.

[50] IEEE P1609.4/D6.0. Draft Standard for Wireless Access in Vehicular Environments (WAVE)

- Multi-channel Operation. Mar. 2010.

[51] L.-W. Chen, Y.-H. Peng, and Y.-C. Tseng. “An Infrastructure-less Framework for Preventing

Rear-End Collisions by Vehicular Sensor Networks.” IEEE Communications Letters, vol. 15,

no. 3, pp. 358-360, Mar. 2011.

[52] QualNet, http://www.scalable-networks.com/products/qualnet/.

[53] FriendlyARM, Mini2440, http://www.friendlyarm.net.

[54] ARM, ARM920T, http://www.arm.com/products/CPUs/ARM920T.html.

[55] H-550EV CO

Sensor Module, http://www.co2sensor.co.kr/new/eng/ndir-co2-

sensor-module-h550ev.htm.

[56] S. Biswas, R. Tatchikou, and F. Dion, “Vehicle-to-Vehicle Wireless Communication

Protocols for Enhancing Highway Traffic Safety,” IEEE Communications Magazine, vol. 44,

no. 1, pp. 74–82, Jan. 2006.

[57] Jennic, JN5139, http://www.jennic.com.

[58] Fastrax, uPatch300, http://www.fastrax.fi.

[60] ITRI WAVE Communication Unit, http://www.itri.org.tw.

[61] IEEE P1609.3/D5.0. Draft Standard for Wireless Access in Vehicular Environments (WAVE)

- Networking Services. Mar. 2010.

VS

3

: A Vehicular Surveillance and Sensing

System for Security Applications

Lien-Wu Chen

_{, Kun-Ze Syue}

_{, and Yu-Chee Tseng}

_{Department of Computer Science, National Chiao-Tung University,}

Hsin-Chu, 300, Taiwan

‡

_{Information and Communications Research Labs, Industrial Technology Research Institute,}

Chu-Tung, Hsin-Chu, 310, Taiwan

Abstract—The Vehicular Surveillance and Sensing System (VS3₎

is a 3G-based mobile device for car security applications. On the car side, it consists of a CO2 sensor, a camera module, a

3G module, and a microprocessor. On the user side, only a 3G mobile phone is needed. VS3 _{provides the following features:}

(i) it can be triggered by events detected on car, (ii) events can be abnormal air quality or potential burglary, and (iii) it supports text or multimedia interaction with users. Application scenarios include detecting an abnormal CO2 level or potential

car burglary, which triggers VS3 _{to transmit SMS, MMS, or}

interactive video call to the vehicle owner, who can then monitor the car situation in return. VS3 _{thus demonstrates a new car}

security and burglarproof prototype.

Keywords: Burglarproof, IEEE 802.15.4, Surveillance, Ve-hicular Sensor Network, Wireless Network.

I. INTRODUCTION

The rapid progress of embedded micro-sensing MEMS and wireless communication technologies has made vehicular

sensor networks (VSNs) possible. A VSN normally consists of

a number of sensors placed on a vehicle to collect environment data and utilizes these sensed data for various purposes. Ex-amples include vehicle tracking, crash prevention, and mobile surveillance [2], [3], [7].

In this work, we are interested in taking advantage of the current 3G or 3.5G mobile systems to enrich user interaction in a VSN. Our goal is to develop a surveillance and sensing system for car security applications. Traditional surveillance systems for vehicle protection rely on roadside cameras for video recording. There are two problems associated with such solutions. First, it requires huge efforts to distinguish targets from many other candidates. Second, since targets are not predefined, the recorded images are usually not clear enough. Further, the volume of videos could be huge, thus requiring a lot of labors.

We propose a 3G-enhanced VSN called vehicular

surveil-lance and sensing system (VS3_{). Only a 3G mobile phone} is needed on the user side, whereas an integrated device with a CO2 sensor, a camera module, a 3G module, and a

Y.-C. Tseng’s research is co-sponsored by MoE ATU Plan, by NSC grants 96-2218-E-009-004, 97-3114-E-009-001, 97-2221-E-009-142-MY3, and 98-2219-E-009-005, by MOEA 98-EC-17-A-02-S2-0048 and 98-EC-17-A-19-S2-0052, and by ITRI, Taiwan.

Fig. 1. System architecture of VS3_.

microprocessor is need on the car side. The microprocessor is responsible for issuing commands and coordinating with other modules. VS3 _{provides the following features: (i) it} can be triggered by events detected on car, (ii) events can be abnormal air quality or potential burglary, and (iii) it supports text or multimedia interaction with users. Only when an event is dectected, the camera module is activated to capture images or record videos of that event. Thus, VS3 _can avoid recording unnecessary videos when nothing happens and improving image/video quality. Application scenarios include detecting an abnormal CO2 level or potential car burglary, which triggers VS3 _{to transmit SMS (short message service),} MMS (multimedia message service), or interactive video call to the vehicle owner, who can then monitor the car situation in return. VS3 _{thus demonstrates a new car security and} burglarproof prototype.

II. SYSTEMARCHITECTURE

Fig. 1 shows the VS3 _{architecture. On the car side, it} consists of a CO2 sensor, a camera module, a 3G module, and a microprocessor. On the user side, only a 3G mobile phone is needed. To illustrate how VS3_{works, we demonstrate} a car security and a car burglarproof applications below. In

Fig. 2. A car security application.

Fig. 3. A car burglarproof application.

the car security application in Fig. 2, after the driver parks the vehicle and activates the car unit, VS3 _{will continuously} check the CO2 concentration in the vehicle for a predefined period. During this period, when it is found that the CO2 concentration is beyond a dangerous threshold, VS3 _{will send} a short message to notify the predefined phone number (user unit). On receipt of the warning message, the owner can return a command short message to VS3_{. According to the command,} VS3_{activates the camera module and initiates a video call to} the owner. Through the live video call, the user can monitor possible abnormal events (such as baby or animal forgotten in the vehicle) by his/her 3G phone. Therefore, lives can be saved in time by the help of VS3_.

In the car burglarproof application in Fig. 3, VS3 _notifies the owner as a potential burglar event is detected (such as door open). Since an immediate action is needed, VS3 _{will directly} record a video clip and send it to the owner via MMS. More importantly, the video clip is a critical clue and evidence to catch the thief. Fig. 4 shows the VS3 _flowchart.

Fig. 4. Flowchart of VS3_.

Fig. 5. Building blocks of the car unit.

Fig. 7. H-550EV CO2 sensor integrated with JN5139.

Fig. 8. Wavecom Q2403A module.

III. IMPLEMENTATIONDETAILS

Fig. 5 shows the building blocks of the car unit. The microprocessor in the car unit is an ARM9 board (Mini2440 [4]) with a 3.5” TFT LCD as shown in Fig. 6, which has a 400MHz 32-bit RISC integer processor (ARM920T [1]), 64MB SDRAM, 64MB Nand Flash, 2MB Nor Flash with BIOS, three serial ports, and a 10/100M Ethernet RJ-45. In particular, Mini2440 can run embedded Linux and WinCE to develop diverse applications.

The CO2module has an H-550EV CO2sensor [5] integrated with Jennic JN5139 [6], which is mounted to Mini2440 via an UART interface. Our prototype is shown in Fig. 7. The CO2 sensor module has 0∼5,000ppm measurement range and

±30ppm accuracy. JN5139 has a 16MIPs 32-bit RISC

proces-sor, a 2.4GHz IEEE 802.15.4-compliant transceiver, 192kB of ROM, and 96kB of RAM. In particular, JN5139 allows the flexibility of supporting mesh networking and packet routing inside a vehicle.

The 3G module is currently implemented by a Wavecom Q2403A GSM/GPRS/CDMA module as shown in Fig. 8, which is controlled by Mini2440 via AT commands. It

per-Fig. 9. CAM130 camera module.

forms SMS, MMS, and video calls as instructed by the ARM9 board.

The camera module is implemented by CAM130 as shown in Fig. 9. It is a CMOS optical sensor. Mini2440 can send a snapshot (record) command to CAM130. In return, a full-resolution, single-frame still picture (video) will be transferred to Mini2440 through the serial port.

In the CO2monitoring application, the concentration thresh-old is set to 1500ppm. We use AT commands to trigger the 3G module to send short messages. The car owner can return a short message with a specific command to ask the ARM9 board to initiate a video call back.

In the burglarproof application, besides a warning short message, a video clip is sent to the car owner as a multimedia message. The clip can be provided to polices as evidence in the future when needed.

IV. CONCLUSION

VS3 _{integrates 3G communication and CO}

2 sensing into surveillance technologies to support intelligent car security applications. The vehicle owner can be informed immediately as unusual events are detected on car. At the same time, the owner can remotely monitor the situation inside vehicle and then take proper actions if necessary. VS3_{can prevent vehicles} form burglar or keep evidences to catch the thief. Furthermore, The baby or animal forgetfully left in the vehicle can be rescued in time by the assistance of VS3_{. The future extension} of VS3 _{could be equipped more various sensors and form a} VS3 _{network to investigate cooperation issues and develop} novel applications.

REFERENCES

[1] ARM, ARM920T. http://www.arm.com/products/CPUs/ARM920T.html. [2] C. Sharp, S. Schaffert, A. Woo, N. Sastry, C. Karlof, S. Sastry, and D.

Culler. Design and implementation of a sensor network system for vehicle tracking and autonomous interception. In Proceeedings of the Second

European Workshop on Wireless Sensor Networks, pages 93–107, 2005.

[3] D. Djenouri. Preventing vehicle crashes through a wireless vehicular sensor network. In 24th Biennial Symposium on Communications, pages 320–323, June 2008.

[4] FriendlyARM, Mini2440. http://www.friendlyarm.net.

[5] H-550EV CO2 Sensor Module. http://www.co2sensor.co.kr/new/eng/ndir-co2-sensor-module-h550ev.htm.

[6] Jennic, JN5139. http://www.jennic.com.

[7] Y.-C. Tseng, Y.-C. Wang, K.-Y. Cheng, and Y.-Y. Hsieh. iMouse:

An integrated mobile surveillance and wireless sensor system. IEEE

A Vehicular Wireless Sensor Network

for CO

₂

Monitoring

Shu-Chiung Hu

, You-Chiun Wang

, Chiuan-Yu Huang

, and Yu-Chee Tseng

1

_{Department of Computer Science, National Chiao-Tung University, Hsin-Chu, 300, Taiwan}

2

_{Department of Information and Computer Engineering, Chung-Yuan Christian University, Chung-Li, 320, Taiwan}

Email:

{schu, wangyc, chiuanyu, yctseng}@cs.nctu.edu.tw

Abstract— Micro-climate monitoring usually requires deploy-ing a large number of measurement tools. By adoptdeploy-ing vehicular wireless sensor networks (VSNs), we can use fewer tools to achieve fine-grained monitoring. This work proposes a VSN architecture to realize micro-climate monitoring based on GSM short messages and availability of GPS receivers on vehicles. We demonstrate our prototype of a ZigBee-based car network

to monitor the concentration of carbon dioxide (CO₂) gas in

areas of interest. The reported data are sent to a server, which is integrated with Google Maps as our user interface. Since mobility of these vehicles is not controllable and sending short messages incurs charges, we also design an on-demand approach to adjust vehicles’ reporting rates to balance between the micro-climate accuracy and the communication cost.

I. INTRODUCTION

We are interested in monitoring micro-climate, which means fine-grained environmental data in the scale of tens to hundreds of square meters. Typically, climate means macro-climate, which means data in the scale of tens to hundreds of square kilometers. Monitoring micro-climate requires a large number of measurement tools. By adopting vehicles (e.g., taxis and buses) as carriers with sensing devices and wireless commu-nication interfaces, we can use fewer measurement tools to achieve fine-grained monitoring. We refer to such systems as

vehicular wireless sensor networks (VSNs).

This paper proposes a VSN architecture to monitor micro-climate based on GSM short messages and geographic infor-mation of vehicles. We show our prototype to monitor the concentration of carbon dioxide (CO₂) gas in areas of interest. CO₂ gas is a critical index of air quality and global warming. In our prototype, a vehicle is equipped with a CO₂ sensor, a GPS receiver, and a GSM module, which form a ZigBee-based intra-vehicle wireless network. Each of such vehicles thus serves as a vehicular sensor. These vehicular sensors roam inside the area of interest and periodically report their sensed data through GSM short messages. The reported data is collected by a server, which is integrated with Google Maps [1] to demonstrate the result.

Since the mobility of these vehicles is not controllable and sending short messages incurs charges, how to adjust vehicles’ reporting rates to balance between the monitoring accuracy and the communication cost is a challenge issue. We propose an adaptive approach to dynamically change the reporting rates

of vehicular sensors on their readings. In particular, the data variation in a grid is considered to adjust the reporting rate.

The major contributions of this paper are two-fold. First, we propose a new architecture based on VSNs to support fine-grained micro-climate monitoring by using a small number of measurement tools. A prototype is also implemented to verify the practicability of the proposed architecture. Second, based on the proposed architecture, we also design an adaptive approach to adjust the reporting rates of vehicles to balance monitoring quality and communication cost.

The rest of this paper is organized as follows. Section II surveys some related work. Section III presents the proposed VSN architecture. Our prototyping experiences are given in Section IV. Section V concludes this paper.

II. RELATEDWORK

Wireless sensor networks have been widely applied to surveillance or monitoring scenarios [2][3][4]. However, they do not discuss how to exploit mobility to reduce monitoring cost. Mobile sensor deployment and dispatch have been in-tensively studied in [5]. BikeNet [6] deploys multiple types of sensors on bicycles to analyze various road information for sharing of cyclists’ experience. MobEyes [7] adopts cameras and chemical sensors to monitor pollution on streets, and vehicles may exchange their sensing data when they meet with each other. Compared to these work, our work is unique in trying to reach a balance between message overheads and sensing quality, under dynamically changing environments.

III. THEPROPOSEDVSN ARCHITECTURE

Fig. 1 illustrates the proposed VSN architecture for micro-climate monitoring. It contains a monitoring server, several vehicular sensors, and GSM networks. Each vehicular sensor is equipped with a CO₂ sensor, a GSM module, and a GPS receiver and periodically reports its sensed CO₂concentration and its current location to the server through GSM short mes-sages. The monitoring server then calculates the distribution of CO₂concentration and renders the result on Google Maps. According to the observed distribution and the vehicle density, the server will ask sensors to adjust their reporting rates. For each vehicular sensor, the intra-vehicle network is a ZigBee network.

GSM networks monitoring server non-sensing vehicle sensing vehicle GSM base station wireless link GSM module CO2 sensor Jennic SMS GPS receiver Jennic

Fig. 1. The proposed VSN architecture for micro-climate monitoring.

We adopt GSM short message service since it is a mature technology. It can be easily extended to 3G or 3.5G technolo-gies. Since sending short messages incurs charges, we need an adaptive approach to adjust sensors’ reporting rates. The basic idea is to partition the monitoring area into grids. Each grid has its own reporting rate according to the variance of CO2 concentration and vehicle density in that grid. Let rmaxk andrkminbe the maximum and minimum CO2 readings of the

k-th grid, respectively, and P_kmax and P_kmin be the positions in gridk where these two readings are reported, respectively.

We define the variance of gird k is as ρk= r

max

k − rkmin

d(P_kmax, P_kmin), (1)

where d(Pkmax, Pkmin) is the distance between Pkmax and Pkmin. Intuitively, ρk indicates how drastic the change of readings is. The number of vehicular sensors in gridk can be

estimated by δk = _µn_kk_×t, where nk is the number of sensing reports received in gridk during an observation interval t and μk is the current reporting rate in gridk.

Intuitively, a higher reporting rateμkshould be set when the variance ρk is higher, and vice versa. For example, in Fig. 2, the variances in grids (2, 4) and (2, 5) are more significant, so higher reporting rates are required. Since grid (2, 5) has more vehicles, its rate can be slightly lower than that of grid (2, 4). Similarly, the variances in grids (1, 5) and (2, 6) are less significant, so lower reporting rates should be adopted to reduce messages. Since grid (2, 6) has more vehicles, its rate can be slightly lower than that of grid (1, 5).

Based on the above observation, our adaptive approach works as follows. Assume that each round is of length t

minutes. Let μmax andμmin be the maximum and minimum allowable reporting rates, respectively. Consider round i. Let ρi_k,δi_k, andμi_k be the variance, the estimated number of

vehi-1 2 3 4 1 2 3 4 5 6 550 ppm 500 ppm 450 ppm CO2 (1,5) (2,4) (2,5) (2,6) concentration

Fig. 2. An example of grid architecture and reporting rate adjustment.

cles, and the reporting rate at round i in grid k, respectively.

We propose to compute the reporting rate μi+1_k based on the observed results in roundsi−1 and i. Specifically, we compute μi+1_k at the beginning of roundi + 1 as follows:

μi+1_k =



min{μmax, tmp} if tmp > μik

max{μmin, tmp} otherwise , where (2)

tmp =  ρi_k ρi−1_k × δi−1_k δ_ki  × μi k. (3)

The value ofμi+1_k should be sent to vehicles at the beginning of roundi + 1.

GSM module CO2sensor GPS receiver inside vehicle GPS receiver GSM module outside vehicle CO2sensor Jennic board

Fig. 3. The snapshots of hardware components.

IV. PROTOTYPINGEXPERIENCES

We have implemented a 16-vehicle prototype to collect CO₂ concentration in Hsin-Chu Science Park, Taiwan. Each vehicle is equipped with the following hardware components (as shown in Fig. 3):

1) Jennic board: It is a microprocess with a wireless module. A Jennic board contains a JN5139 chip [8], which has a 32-bit RISC processor, a fully compliant 2.4 GHz IEEE 802.15.4 [9] transceiver, 192 KB of ROM, and 96 KB of RAM. We use the ZigBee protocol [10] for inter-board communication.

2) GPS receiver: We adopt the uPatch300 GPS module [11]. It can provide geographic location with accuracy

≤ 1.8 meters. Its reporting rate is set to 1 second.

3) CO₂sensor: We adopt the H-550EV CO₂sensor module [12]. It will sample CO₂ concentration every 3 seconds. Its detectable range is from 0 to 5,000 ppm with error range of±30 ppm.

4) GSM module: We adopt the SIM300 GSM module [13], which supports the tri-band GSM/GPRS communica-tion on frequency bands of 900 MHz, 1,800 MHz, and 1,900 MHz.

Fig. 3 shows the snapshots of these components. The CO2 sensor is installed outside the vehicle, while the GPS receiver and the GSM module are installed inside the vehicle. Each of the GPS receiver and the CO₂ sensor is attached to a Jennic board, so they can communicate with each other through a ZigBee wireless link. The GPS receiver is connected to the GSM module through an RS232 wired interface. The CO₂ sensor reports its readings periodically at a fixed rate to Jennic board inside the vehicle. The Jennic board will then average these readings, combine them with the current location of the vehicle, and report to the monitoring server via GSM short messages. The reporting will follow the requested rate.

CO2 density: 429 ppm

380 389 390 399 400 409 410 419 420 429

CO2(ppm)

Fig. 4. A snapshot of CO₂concentration at the NCTU campus.

Fig. 4 demonstrates our monitoring results at the National Chiao-Tung University (NCTU) campus. The monitoring re-gion is approximately 80 hectares and is partitioned into 5 × 4 grids. The observed CO2 concentration ranges from 380 ppm to 429 ppm. Each circle indicates the monitoring position and its color represents the corresponding level of CO₂ concentration. Users can click on each circle to obtain the detailed data.

V. CONCLUSIONS ANDFUTUREWORK

In this paper, we have proposed a new architecture based on VSNs for micro-climate monitoring. Through GSM short messages and geographic locations of vehicles, we can use a small number of vehicles to realize a fine-grained monitoring in urban areas. To balance between the monitoring quality and the message cost, we have designed an adaptive approach to adjust the reporting rates of sensing vehicles according to the variance of sensing readings and the density of vehicles

in each grid. We have also demonstrated the prototype of a ZigBee-based intra-vehicle wireless network.

ACKNOWLEDGEMENT

Y.-C. Tseng’s research is co-sponsored by MoE ATU Plan, by NSC grants 96-2218-E-009-004, 3114-E-009-001, 97-2221-E-009-142-MY3, and 98-2219-E-009-005, and by ITRI, Taiwan.

REFERENCES

[1] Google Maps. [Online]. Available: http://maps.google.com/

[2] T. He, S. Krishnamurthy, J. A. Stankovic, T. Abdelzaher, L. Luo, R. Stoleru, T. Yan, L. Gu, G. Zhou, J. Hui, and B. Krogh, “VigilNet: an integrated sensor network system for energy-efficient surveillance,”

ACM Trans. on Sensor Networks, vol. 2, no. 1, pp. 1–38, 2006.

[3] L. E. Cordova-Lopez, A. Mason, and J. D. Cullen, “Online vehicle and atmospheric pollution monitoring using gis and wireless sensor networks,” in Proc. of ACM Int’l Conference on Embedded Networked

Sensor Systems (SenSys), 2007, pp. 87–101.

[4] K. Liu, M. Li, Y. Liu, M. Li, Z. Guo, and F. Hong, “Passive diagnosis for wireless sensor networks,” in Proc. of ACM Int’l Conference on

Embedded Networked Sensor Systems (SenSys), 2008, pp. 113–126.

[5] Y.-C. Wang, F.-J. Wu, and Y.-C. Tseng, “Mobility management algo-rithms and applications for mobile sensor networks,,” Wireless

Commu-nunications and Mobile Computing (WCMC), to appear.

[6] S. B. Eisenman, E. Miluzzo, N. D. Lane, R. A. Peterson, G. S. Ahn, and A. T. Campbell, “The BikeNet mobile sensing system for cyclist experience mapping,” in Proc. of ACM Int’l Conference on Embedded

Networked Sensor Systems (SenSys), 2007, pp. 87–101.

[7] U. Lee, B. Zhou, M. Gerla, E. Magistretti, P. Bellavista, and A. Corradi, “Mobeyes: smart mobs for urban monitoring with a vehicular sensor network,” IEEE Wireless Communications, vol. 13, no. 5, pp. 52–57, 2006.

[8] Jennic JN5139. [Online]. Available: http://www.jennic.com/

[9] “IEEE standard for information technology - telecommunications and information exchange between systems - local and metropolitan area networks specific requirements part 15.4: wireless medium access con-trol (MAC) and physical layer (PHY) specifications for low-rate wireless personal area networks (LR-WPANs)(revision of IEEE Std 802.15.4-2003),” 2006.

[10] “ZigBee specification version 2006, ZigBee document 064112,” 2006. [11] uPatch300 module. [Online]. Available: http://www.fastraxgps.com/ [12] H-550EV module. [Online]. Available: http://www.elti.co.kr/ [13] SIM300 module. [Online]. Available: http://www.sim.com/