陳薈慧+郭斌+於志文
Mobile Crowd-Sensing Application
中圖分類號:TP391 文獻標志碼:A 文章編號:1009-6868 (2014) 01-0035-003
摘要:認為無線通信和智能移動設備的發(fā)展為群智感知在移動環(huán)境下的應用奠定了基礎,而廉價多樣的傳感器使移動群智感知應用與人類社會的聯(lián)系更加緊密。移動群智感知用戶采集數(shù)據(jù)時的協(xié)作方式分為參與式感知、機會感知兩種,各有優(yōu)缺點和局限性。移動群智感知需要考慮用戶成本、網(wǎng)絡壓力、云計算服務器架設、用戶隱私保護等方面的問題,要面對情境隱私、匿名任務、匿名數(shù)據(jù)匯報、可靠數(shù)據(jù)讀取、數(shù)據(jù)真實性等安全方面的挑戰(zhàn)。
關鍵詞: 移動群智感知;參與式感知;機會感知
Abstract: The development of wireless communication and smart mobile devices has been the impetus for mobile crowd-sensing applications. Low-cost sensors in smart devices means that mobile crowd-sensing applications are more tightly associated with human communities. In a mobile crowd-sensing application, measures for sensing cooperation between individuals may be categorized as participant-sensing or opportunity-sensing. Both of these measures have advantages and disadvantages. Mobile crowd-sensing has to consider problems such as cost to the user, pressure on the mobile communication network, constructing a cloud server, and user privacy. Security is a challenge in privacy protection, anonymous tasking, anonymous reporting, collection of dependent data, and data reliability.
Key words: mobile crowd-sensing application; participant sensing; opportunity sensing
1 群智感知的架構
群智感知通過感知個體的信息而挖掘群體信息并反作用于個體或群體[1]。群智感知是個體與群體的合作與共贏,主體是“感知”和“挖掘”,感知層由個體與攜帶的智能設備組成,挖掘層由后臺數(shù)據(jù)服務器組成。隨著數(shù)據(jù)量的爆炸式增長,云計算開始為數(shù)據(jù)存儲和挖掘提供支持,傳感器和應用程序完成數(shù)據(jù)的采集與群體感知結果的反饋,如圖1所示。
感知層完成數(shù)據(jù)的采集,無論是參與式感知還是機會感知,都由終端采集用戶數(shù)據(jù)并上傳。挖掘層通常是為了發(fā)現(xiàn)某種知識或者統(tǒng)計某種結果而對大數(shù)據(jù)進行深層分析。
2 群智感知的數(shù)據(jù)采集
移動群智感知采集的數(shù)據(jù)不再僅僅局限于位置,移動設備附帶的各種傳感器在個體數(shù)據(jù)采集時都能夠發(fā)揮作用。例如,路人通過分析手機麥克風采集到的環(huán)境聲音檢測環(huán)境噪聲,旅行者通過手機攝像頭和GPS記錄旅游日志并分享旅游攻略,晨練者通過加速度傳感器監(jiān)測運動量并結合GPS軌跡分享晨練感受,司機或者乘客通過加速度傳感器采集道路坑洼狀況上傳給城市管理部門。
根據(jù)手機用戶采集數(shù)據(jù)時的協(xié)作方式,可將感知分為參與式感知[2]和機會感知[3]。參與式感知由用戶主動參與,因此數(shù)據(jù)精度高但容易受用戶主觀意識干擾。機會感知通過直接或間接方式感知用戶的行為,對用戶干擾較小,但數(shù)據(jù)精度依賴于感知算法和應用環(huán)境,且機會感知需較高的隱私保護機制激勵用戶的參與。
參與式感知的實時性相對不如機會感知高,但機會感知準確采集數(shù)據(jù)的難度要比參與式感知高。比如,在交通路況監(jiān)控應用中,參與式感知需要用戶主動上傳數(shù)據(jù),但是當司機上報擁堵信息時可能已經(jīng)離開了該路段,導致時空信息不準確,另外,要求司機不停的上傳數(shù)據(jù)的可能性較小。機會感知雖然不要求司機手動提交數(shù)據(jù),但是對路況的準確感知卻是亟待解決的問題。
參與式感知采集到的數(shù)據(jù)更容易受到用戶的主觀干擾,應用系統(tǒng)需要對數(shù)據(jù)進行有效性的判斷并向用戶提供量化的評級標準,例如,Creek Watch應用向用戶提供水質判斷標準;Talasila等人[4]使用定位信息判斷用戶提交的交通堵塞信息是否真實,防止有人故意在出發(fā)前發(fā)布假的擁堵信息用于清暢道路;Wreckwatch[5]應用使用車載手機的加速度計感知到撞車以后判斷乘客的受傷程度,使用VoIP電話、短消息、GPS地圖和路人拍照上報等方法對撞車情況和人員受傷情況進一步核實。
3 群智感知的應用現(xiàn)狀
現(xiàn)階段的移動群智感知應用大致可分為3類:環(huán)境、公共設施和社會[6]。
在環(huán)境方面的應用如Common Sense[7]、Creek Watch[8-9]、Ear-Phone[10]和iMap[11]等。Common Sense使用可以與手機通信的手持空氣質量傳感器收集空氣污染數(shù)據(jù)(如二氧化碳、氮氧化物),分析和可視化后通過Web發(fā)布;Creek Watch是由IBM在2010年11月發(fā)布的iPhone應用,人們路過河流的時候,可以花費幾秒鐘的時間搜集水質數(shù)據(jù),包括流量、流速和垃圾數(shù)量,后臺服務器匯總數(shù)據(jù)后在網(wǎng)站上公布;Ear-Phone使用手機根據(jù)噪音級別監(jiān)測對人類聽力有害的噪音污染,并繪制成噪音地圖通過Web共享;iMap使用手機采集人的時間-地點軌跡,并使用已有模型計算空氣中二氧化碳的含量和PM2.5的值,實現(xiàn)間接環(huán)境監(jiān)測功能。
在公共設施方面的應用如交通擁堵情況的檢測[12-13]、道路狀況的檢測[14](如道路坑洼、噪音)、尋找停車位[15]、公共設施報修(如消防栓、交通信號燈、井蓋等)和實時交通監(jiān)測與導航[16]等。例如,ParkNet使用GPS和安裝在右側車門的超聲波傳感器檢測空停車位,并共享檢測結果;CMS系統(tǒng)收集由公交車乘客的手機采集的數(shù)據(jù),對公交車舒適程度做出評級,并通過網(wǎng)站發(fā)布;Zhou P等人[17]設計了Android平臺下的公交車到站時刻預測系統(tǒng);GBus[18]應用允許個人使用移動設備收集公交車站點信息,包括站點名稱、圖片和描述;EasyTracker[19]應用使用安裝有地圖的智能手機,從GPS軌跡中提取高密度點獲取公交站點,并采集各站點公交到站時刻計算公交站點間運行時間,從而預測公交到站時刻。
在社會方面的應用如社交網(wǎng)絡應用以及社會感知。例如,騰訊提供的根據(jù)個體之間的共同好友而進行的好友推薦機制;Ubigreen[20]應用通過手機感知和用戶參與的形式半自動采集用戶出行習慣,鼓勵用戶綠色出行;im2GPS[21]應用構建自己的GPS照片知識庫,使人們可以通過拍攝照片查詢自己所處的位置;DietSense[22]應用允許用戶在社交群中分享個人飲食習慣,人們可以比較自己的飲食習慣并向他人提出建議;Bikenet[23]應用根據(jù)個體提供的自行車騎行路線的GPS軌跡、空氣質量、噪聲質量等數(shù)據(jù)計算出最適合自行車騎行運動的路線。
4 群智感知面臨的問題與
挑戰(zhàn)
移動群智感知可以通過個體數(shù)據(jù)完成大規(guī)?,F(xiàn)象的監(jiān)測[24],因此需要收集大量來自個體的數(shù)據(jù)。為了使計算過程更加高效,通常需要考慮如下幾個問題:
·數(shù)據(jù)的收集必須考慮用戶成本
·用戶傳輸數(shù)據(jù)過程中對網(wǎng)絡造成的壓力
·需要架設用于接收、計算、管理和分析的后臺云計算服務器
·使用戶能夠放心并自愿將手機作為數(shù)據(jù)收集探測器的用戶隱私保護機制
Kapadia A等人提出了機會感知環(huán)境下安全的9個挑戰(zhàn):
·情境隱私
·匿名任務
·匿名數(shù)據(jù)匯報
·可靠數(shù)據(jù)的讀取
·數(shù)據(jù)真實性
·系統(tǒng)集成
·防止數(shù)據(jù)抑制
·參與
·公平
群智環(huán)境下數(shù)據(jù)可信度問題包括系統(tǒng)對個體采集端數(shù)據(jù)的驗證和系統(tǒng)對事件結果的準確判斷。采集數(shù)據(jù)的可靠性一方面需要采用信用規(guī)范對客戶端用戶加以限制,另一方面需要更精確的算法來驗證數(shù)據(jù)的可靠性。系統(tǒng)對事件結果的準確判斷是激勵用戶繼續(xù)使用該系統(tǒng)的關鍵基礎。群智環(huán)境雖然為假數(shù)據(jù)提供了上傳的機會,但真實數(shù)據(jù)也擁有相同的機會,服務器如何分辨這些數(shù)據(jù)的真?zhèn)问且淮筇魬?zhàn)。
智能手機應用程序的隱私保護機制是用戶普遍關心的問題。如果沒有隱私保護,以用戶為中心的機會感知永遠不可能成功。在數(shù)據(jù)上傳時加入干擾數(shù)據(jù)或者對敏感數(shù)據(jù)加密是常用的保護機制。
從數(shù)據(jù)采集角度來講,手機附帶的傳感器可以透露用戶所處的環(huán)境,并能實時記錄語音和圖像,這些功能如果應用不當就容易演變成非法侵犯個人隱私的應用,如果用戶采用過高的安全保護策略,則移動群智感知無法得到豐富的終端數(shù)據(jù)。
移動群智應用環(huán)境下,用戶不可避免地需要上傳位置、手機號碼等信息,而手機號碼等個人信息可以與用戶本人綁定,所以,用戶有理由懷疑應用程序可能會侵犯個人隱私。如果將個人數(shù)據(jù)加密或者保證使用更高安全級別的服務器,但仍然不能保證用戶不懷疑數(shù)據(jù)丟失時隱私受到侵犯,這與困擾著云服務的隱私問題是一樣的。
3G、4G、Wi-Fi以及未來的通信技術能夠為群智感知應用解決帶寬的問題,但移動群智應用系統(tǒng)與其他移動應用軟件一樣還存在能耗和數(shù)據(jù)流量的問題?,F(xiàn)有的節(jié)電方法包括選用合適的數(shù)據(jù)采集觸發(fā)算法和選用低能耗的傳感器等,控制通信流量的方法主要采用有選擇的上傳高質量的數(shù)據(jù)。
5 結束語
“眾人拾柴火焰高”,對于現(xiàn)代人類社會來說,移動群智感知是群體智慧在信息時代的表現(xiàn)方式。無線網(wǎng)絡和智能設備是移動群智感知的基礎,移動群智感知應用需求的不斷擴展反過來對這兩者的發(fā)展又提出了更高的要求。群智感知環(huán)境下的個體以數(shù)字的形式存在,隱私保護是個體互信協(xié)作的基礎,安全規(guī)范和信用規(guī)范的不完善勢必影響群智感知應用推廣??偟膩碚f,移動群智感知對人類社會的發(fā)展和信息科技的發(fā)展都將起到促進的作用,而更多的挑戰(zhàn)也等待著人們?nèi)グl(fā)現(xiàn)和解決。
參考文獻
[1] 劉云浩. 群智感知計算 [J]. 中國計算機學會通訊, 2012,8(10):38-42.
[2] KANHERE S S. Participatory sensing: Crowdsourcing data from mobile smartphones in urban spaces [C]//Proceedings of the International Conference on Distributed Computing and Internet Technology, ICDCIT 2013, Springer Berlin Heidelberg, LNCS 7753, 2013: 19-26.
[3] KAPADIA A, KOTZ D, TRIANDOPOULOS N. Opportunistic sensing: Security challenges for the new paradigm [C]//Proceedings of the IEEE Communication Systems and Networks and Workshops, COMSNETS 2009, Bangalore, India, 2009: 1-10.
[4] TALASILA M, CURTMOLA R, BORCEA C. Improving location reliability in crowd sensed data with minimal efforts [C]//Proceedings of the 6th Joint IFIP/IEEE Wireless and Mobile Networking Conference, WMNC, 2013.
[5] WHITE J, THOMPSON C, TURNER H, et al. WreckWatch: automatic traffic accident detection and notification with smartphones [J]. Mobile Networks and Applications, 2011, ACM, 2011,16(3): 285-303. doi:10.1007/s11036-011-0304-8.
[6] GANTI R K, YE F, LEI H. Mobile crowdsensing: Current state and future challenges [J]. Communications Magazine, IEEE, 2011,49(11): 32-39. doi:10.1109/MCOM.2011.6069707.
[7] DUTTA P, AOKI P M, KUMAR N, et al. Common sense: participatory urban sensing using a network of handheld air quality monitors [C]//Proceedings of the 2009 ACM Conference on Embedded Networked Sensor Systems, SenSys 2009, Berkeley, California, USA, 2009: 349-350.
[8] IBM research. Creek watch - explore your watershed [EB/OL]. (2013-10-10). http://creekwatch.researchlabs.ibm.com/.
[9] KIM S, ROBSON C, ZIMMERMAN T, et al. Creek watch: pairing usefulness and usability for successful citizen science [C]//Proceedings of the 2011 ACM Conference on Human Factors in Computing Systems, CHI 2011, Vancouver, Canada, 2011: 2125-2134.
[10] RANA R K, CHOU C T, KANHERE S S, et al. Ear-phone: an end-to-end participatory urban noise mapping system [C]//Proceedings of the ACM/IEEE International Conference on Information Processing in Sensor Networks, IPSN 2010, Stockholm, Sweden, 2010: 105-116.
[11] DEMIRBAS M, RUDRA C, RUDRA A, et al. iMap: Indirect measurement of air pollution with cellphones [C]//Proceedings of the IEEE Conference of Pervasive Computing and Communications, PerCom 2009, Galveston, Texas, USA, 2009: 1-6.
[12] LEONTIADIS I, MARFIA G, MACK D, et al. On the effectiveness of an opportunistic traffic management system for vehicular networks [J]. Intelligent Transportation Systems, IEEE Transactions on, 2011, 12(4): 1537-1548. doi: 10.1109/TITS.2011.2161469.
[13] LIU S, LIU Y, NI L, et al. Detecting crowdedness spot in city transportation [J]. Vehicular technology, IEEE Transactions on, 2013, 62(4):1527-1539. doi: 10.1109/TVT.2012.2231973.
[14] LIN C Y, CHEN L J, CHEN Y Y, et al. A comfort measuring system for public transportation systems using participatory phone sensing [C]//Proceedings of the ACM PhoneSense, 2010, Zurich, Switzerland, 2010.
[15] MATHUR S, JIN T, KASTURIRANGAN N, et al. Parknet: Drive-by sensing of road-side parking statistics [C]//Proceedings of the ACM International Conference on Mobile Systems, applications, and services, MobiSys 2010, San Francisco, CA, USA, 2010: 123-136.
[16] MANOLOPOULOS V, TAO S, RODRIGUEZ S, et al. MobiTraS: A mobile application for a Smart Traffic System [C]//Proceedings of the IEEE Conference NEWCAS, 2010, Montreal, Canada, 2010: 365-368.
[17] ZHOU P, ZHENG Y, LI M. How long to wait?: predicting bus arrival time with mobile phone based participatory sensing [C]//Proceedings of the ACM Conference on Mobile systems, applications, and services. MobiSys 2012, Low Wood Bay, UK, 2012: 379-392.
[18] SANTOS M, PEREIRA R L, LEAL A B. GBUS-Route GeoTracer [C]//Proceedings of the Workshop on Vehicular Traffic Management for Smart Cities, VTM 2012, IEEE, 2012: 1-6.
[19] BIAGIONI J, GERLICH T, MERRIFIELD T, et al. Easytracker: Automatic transit tracking, mapping, and arrival time prediction using smartphones [C]//Proceedings of the ACM Conference on Embedded Networked Sensor Systems. SenSys2011, Seattle, WA, USA, 2011: 68-81.
[20] FROEHLICH J, DILLAHUNT T, KLASNJA P, et al. UbiGreen: Investigating a mobile tool for tracking and supporting green transportation habits [C]//Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, CHI 2009, Boston, USA , 2009: 1043-1052.
[21] JAMES H, ALEXEI E. IM2GPS: Estimating geographic information from a single image [C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2008, Anchorage, AK, USA, 2008:1-8.
[22] REDDY S, PARKER A, HYMAN J, et al. Image browsing, processing, and clustering for participatory sensing: lessons from a DietSense prototype [C]//Proceedings of the ACM Workshop on Embedded networked sensors, 2007: 13-17.
[23] EISENMAN S B, MILUZZO E, LANE N D, et al. BikeNet: A mobile sensing system for cyclist experience mapping [J]. ACM Transactions on Sensor Networks, 2009, 6(1): 6. doi: 10.1145/1653760.1653766.
[24] SHERCHAN W, JAYARAMAN P P, KRISHNASWAMY S, et al. Using on-the-move mining for Mobile crowdsensing [C]//Proceedings of the IEEE Conference on Mobile Data Management, MDM 2012, Bengaluru, Karnataka, 2012: 115-124.
作者簡介
陳薈慧,西北工業(yè)大學在讀博士研究生,洛陽理工學院講師;主要研究方向為普適計算;已發(fā)表論文10余篇。
郭斌,西北工業(yè)大學副教授、碩士生導師;主要研究方向為普適計算、群智感知;已發(fā)表論文60余篇,主持基金項目2項。
於志文,西北工業(yè)大學教授、博士生導師;主要研究方向為普適計算、群智感知;已發(fā)表論文100余篇,主持基金項目4項。
[16] MANOLOPOULOS V, TAO S, RODRIGUEZ S, et al. MobiTraS: A mobile application for a Smart Traffic System [C]//Proceedings of the IEEE Conference NEWCAS, 2010, Montreal, Canada, 2010: 365-368.
[17] ZHOU P, ZHENG Y, LI M. How long to wait?: predicting bus arrival time with mobile phone based participatory sensing [C]//Proceedings of the ACM Conference on Mobile systems, applications, and services. MobiSys 2012, Low Wood Bay, UK, 2012: 379-392.
[18] SANTOS M, PEREIRA R L, LEAL A B. GBUS-Route GeoTracer [C]//Proceedings of the Workshop on Vehicular Traffic Management for Smart Cities, VTM 2012, IEEE, 2012: 1-6.
[19] BIAGIONI J, GERLICH T, MERRIFIELD T, et al. Easytracker: Automatic transit tracking, mapping, and arrival time prediction using smartphones [C]//Proceedings of the ACM Conference on Embedded Networked Sensor Systems. SenSys2011, Seattle, WA, USA, 2011: 68-81.
[20] FROEHLICH J, DILLAHUNT T, KLASNJA P, et al. UbiGreen: Investigating a mobile tool for tracking and supporting green transportation habits [C]//Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, CHI 2009, Boston, USA , 2009: 1043-1052.
[21] JAMES H, ALEXEI E. IM2GPS: Estimating geographic information from a single image [C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2008, Anchorage, AK, USA, 2008:1-8.
[22] REDDY S, PARKER A, HYMAN J, et al. Image browsing, processing, and clustering for participatory sensing: lessons from a DietSense prototype [C]//Proceedings of the ACM Workshop on Embedded networked sensors, 2007: 13-17.
[23] EISENMAN S B, MILUZZO E, LANE N D, et al. BikeNet: A mobile sensing system for cyclist experience mapping [J]. ACM Transactions on Sensor Networks, 2009, 6(1): 6. doi: 10.1145/1653760.1653766.
[24] SHERCHAN W, JAYARAMAN P P, KRISHNASWAMY S, et al. Using on-the-move mining for Mobile crowdsensing [C]//Proceedings of the IEEE Conference on Mobile Data Management, MDM 2012, Bengaluru, Karnataka, 2012: 115-124.
作者簡介
陳薈慧,西北工業(yè)大學在讀博士研究生,洛陽理工學院講師;主要研究方向為普適計算;已發(fā)表論文10余篇。
郭斌,西北工業(yè)大學副教授、碩士生導師;主要研究方向為普適計算、群智感知;已發(fā)表論文60余篇,主持基金項目2項。
於志文,西北工業(yè)大學教授、博士生導師;主要研究方向為普適計算、群智感知;已發(fā)表論文100余篇,主持基金項目4項。
[16] MANOLOPOULOS V, TAO S, RODRIGUEZ S, et al. MobiTraS: A mobile application for a Smart Traffic System [C]//Proceedings of the IEEE Conference NEWCAS, 2010, Montreal, Canada, 2010: 365-368.
[17] ZHOU P, ZHENG Y, LI M. How long to wait?: predicting bus arrival time with mobile phone based participatory sensing [C]//Proceedings of the ACM Conference on Mobile systems, applications, and services. MobiSys 2012, Low Wood Bay, UK, 2012: 379-392.
[18] SANTOS M, PEREIRA R L, LEAL A B. GBUS-Route GeoTracer [C]//Proceedings of the Workshop on Vehicular Traffic Management for Smart Cities, VTM 2012, IEEE, 2012: 1-6.
[19] BIAGIONI J, GERLICH T, MERRIFIELD T, et al. Easytracker: Automatic transit tracking, mapping, and arrival time prediction using smartphones [C]//Proceedings of the ACM Conference on Embedded Networked Sensor Systems. SenSys2011, Seattle, WA, USA, 2011: 68-81.
[20] FROEHLICH J, DILLAHUNT T, KLASNJA P, et al. UbiGreen: Investigating a mobile tool for tracking and supporting green transportation habits [C]//Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, CHI 2009, Boston, USA , 2009: 1043-1052.
[21] JAMES H, ALEXEI E. IM2GPS: Estimating geographic information from a single image [C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2008, Anchorage, AK, USA, 2008:1-8.
[22] REDDY S, PARKER A, HYMAN J, et al. Image browsing, processing, and clustering for participatory sensing: lessons from a DietSense prototype [C]//Proceedings of the ACM Workshop on Embedded networked sensors, 2007: 13-17.
[23] EISENMAN S B, MILUZZO E, LANE N D, et al. BikeNet: A mobile sensing system for cyclist experience mapping [J]. ACM Transactions on Sensor Networks, 2009, 6(1): 6. doi: 10.1145/1653760.1653766.
[24] SHERCHAN W, JAYARAMAN P P, KRISHNASWAMY S, et al. Using on-the-move mining for Mobile crowdsensing [C]//Proceedings of the IEEE Conference on Mobile Data Management, MDM 2012, Bengaluru, Karnataka, 2012: 115-124.
作者簡介
陳薈慧,西北工業(yè)大學在讀博士研究生,洛陽理工學院講師;主要研究方向為普適計算;已發(fā)表論文10余篇。
郭斌,西北工業(yè)大學副教授、碩士生導師;主要研究方向為普適計算、群智感知;已發(fā)表論文60余篇,主持基金項目2項。
於志文,西北工業(yè)大學教授、博士生導師;主要研究方向為普適計算、群智感知;已發(fā)表論文100余篇,主持基金項目4項。