標題: 基於M2M聯網之雲端與巨量資料運算平台開發
Cloud and Big Data Computing Platforms for M2M Communication Systems
作者: 曾煜棋
Tseng Yu-Chee
國立交通大學資訊工程學系(所)
公開日期: 2013
摘要: 通信網路技術的發展,使人與人之間可以更加快捷地溝通,資訊的交流更加順暢。 但是目前僅各式電腦、手機和其他一些 IT 類設備具備這種通信和網路能力,眾多的一 般機器設備 (machines) 幾乎不具備聯網和通信能力,例如家電、車輛、自動販賣機、 工程設備等。Machine-to-Machine (M2M)物聯網技術即是使所有的機器設備都具備聯 網和通信能力,達到「一切聯網」(Network Everything)的目標。美國研究機構 Forrester 預測,M2M 技術所帶來的產業價值,會比互聯網 (Internet) 大 30 倍,將形成下一個兆 元級的資通訊產業。M2M 技術的實現和產業推廣,同時需要雲端計算(Cloud Computing) 和巨量資料(Big Data)分析核心技術的輔助配合。為因應這個需要,交通大學向國科 會提出成立「網路與資訊系統工業基礎深耕中心」,來執行「M2M 聯網之雲端與巨量資料 運算平台開發計劃」。本中心將從「研究深耕」、「教學紮根」、「產學密切合作」三個主 軸來推動這個計畫。 在「研究深耕」主軸方面,將分成三個子計畫來執行:(1)M2M 聯網,(2)雲端計算技 術,(3)巨量資料技術。三個子計畫將通過共同平台方式來進行整合,M2M 子計畫研發 M2M 通信協定、M2M 閘道、M2M 終端器與 M2M 服務平台等技術,雲端計算子計畫著力於虛 擬主機、雲端系統網路、雲端密文資料處理系統及雲端應用等技術,巨量資料子計畫研 究巨量資料儲存平台、以分散式資料庫為基礎之平行化 Big Data 探勘演算法及物聯網 之物件資料收集與探勘架構等技術,上述項目掌握了要深耕「網路與資訊系統工業基礎」 最重要的共通基礎技術。我們將致力 M2M 相關工業所需基礎,並展示一些整合的 M2M 創 新應用。 在「教學紮根」主軸方面,我們將以研究成果推動教學活動,以傳授學生們關鍵的 知識,以達到「十年磨一劍」和「精益求精」的深耕效果。本中心在教學方面,將推動 以下事項: ● 維基夥伴獎學金及實驗室:提供獎學金給予大學部學生進入維基夥伴實驗室, 及早加入研究及專題製作,並與老師建立長期師徒關係。 ● 學分學程:本中心將與交大資訊學院的二項學程合作(教育部補助之「物聯網』」 學程及台達電子補助之「雲端運算系統」學程),以系列課程紮根教學,擴大 並深耕訓練課程。 ● 產研博士班:為改變以論文發表為目標的現況,交大資訊學院目前正嘗試建立 產研型態博士班,鼓勵學生以產業未來所需的題材為目標,在研究方法及研發 指標方面,都將密切與產業之尖端需求為導向。 在「產學密切合作」主軸方面,本中心將置於交大鑽石計畫辦公室之下,以交大電 子資訊中心的空間和研究員為基礎,鼓勵廠商和學校進行長期且高挑戰的研發工作。本 中心若得以成立,將整合現有的三個團隊研發計畫,加入共同運作:(1)友訊科技鑽石 計畫(已簽約三年,目前執行進入第三年,約滿後將爭取續約)、(2)台達電子(已簽約 三年,目前執行第一年,且正訂定第二年度在雲端運算之研發題目中)、(3)宏達電子(簽 訂 MOU 中,預計自 2013 年起執行)。本中心將以上述產學合作計畫來參與基礎研發中心 之運作,以真正達到深耕網路與資訊系統工業基礎之目的。
The development of communications networks allows people to communicate with each other more rapidly and conveniently. It also enables the exchange of information more smoothly. However, only computers, handsets and some special IT equipment are connected to the communications networks and capable of communicating with each other. Majority of machines, such as home appliances, vehicles, vending machines and industrial equipment, are not equipped with network connectivity. The emerging of M2M technologies will enable all machines with network connectivity and achieve the goal of “network everything”. Forrester, a research firm in the United States, forecasts the value of M2M industry will be 30 times bigger than the value of the Internet and will form the next trillion level of ICT business. The realization of M2M technological vision and the development of its ecosystem, however, require the support of both Cloud Computing and Big Data technologies. In response to this need, we propose to NSC the establishment of a “Deep Plowing Center for the Industrial Base of Network and Information Systems” to work on “Cloud and Big Data Computing Platforms for M2M Communications Systems”. The center will promote this project from three key perspectives: deep plowing research, rooted education and industry-university collaboration. In the perspective of deep plowing research, the project will divide into and execute three subprojects: M2M communications, Cloud Computing and Big Data. The integration of these three subprojects will be done through the development of a common platform. M2M subproject will research and develop technologies of M2M communications protocols, M2M gateways, M2M devices and M2M service platform. Cloud Computing subproject will focus on the technologies of CPU virtualization, Cloud interconnection networks, Cloud security and Cloud applications. Big Data subproject will investigate the technologies of Big Data storage platform, Parallelized Big Data mining based on distributed database and M2M data collection and mining. In the perspective of rooted education, we will drive the teaching activities based on the results of our research so as to achieve the effect of “making a good sword with a decade of the effort” and “building excellence above excellence”. The center will engage the following education activities: • Wiki partnership scholarship and laboratories: The center will provide scholarship for undergraduate students to enter Wiki laboratories so that they can participate in research and special projects earlier and establish long-term disciple relations with professors. • Special credit curriculum: The center will collaborate with the College of Computer Science at NCTU on two special credit curricula (IoT/M2M curriculum sponsored by MOE and Cloud computing system curriculum sponsored by Delta Electronics). We plan to expand and deep plow our training of students based on solid, well-rooted and well-structured curricula. • Industry-oriented doctoral program: The College of Computer Science at NCTU is currently trying to set up an industry-oriented doctoral program to encourage students focus their research on the future needs of industry. The emphasis is to couple doctoral research methods and benchmarks closely with the top-notch needs of industry. In the perspective of industry-university collaboration, our center will be established under the Diamond Program Office (DPO) at NCTU, housed in the building of Microelectronics and Information Systems Research Center (MIRC) at NCTU and supported by the dedicated staff at MIRC. Both DPO and MIRC have been a successful vehicle in encouraging the collaboration between industry and university for long-term challenging projects. With the establishment of the center, we will integrate three industry partners under the operations of the center. These include the ongoing D-link Diamond Project (currently in the third year of a three-year contract and expect the contract to be renewed), the ongoing Delta Electronics program (currently in the first year of a three-year contract with the 2nd year planned to focus on Cloud Computing topics) and the upcoming HTC program (MOU currently being signed, expect the project to start from 2013). The center will incorporate these three industry-university collaborations into its launch and operations to accomplish the objectives of deep plowing the industry base of network and information systems.
官方說明文件#: NSC102-2218-E009-002
URI: http://hdl.handle.net/11536/92196
https://www.grb.gov.tw/search/planDetail?id=2865283&docId=407700
Appears in Collections:Research Plans