ZigBee在無線網路技術中，乃屬較新之無線傳輸技術, 其特色為低電耗、可控制大量感測裝置、反應時間短、建置成本低廉等。其遵循IEEE802.15.4標準與ZigBee通訊協定標準，目前已有ZigBee Alliance制定相關標準與推動此技術，而控制晶片以德州儀器(TI)為主要制訂及生產供應商。ZigBee網路支援星狀、樹狀、網狀三種網路架構，其裝置角色有Coordinator、Router, End Device三種。目前主要應用有軍事為健康、環境、健康、家庭及商業等方面。應用上除了利用韌體方式實作裝置對裝置的控制，有些應用尚需借由應用程式才能達成。本論文主要強調利用API讓應用程式控制裝置取得感測資料更為容易，其亦可解決同時回報或監測所有裝置之問題。 目前整個API針對裝置晶片控制指令集定義成一類別，此部份可依不同晶片版本做類別庫切換。另外，定義如何與Serial Port溝通，以及傳輸指令或資料時是否符合ZigBee通訊協定要求之判斷。基於以上類別庫之定義，最後定義DeviceManager類別,當作管理裝置之物件，而該類別中，為讓開發者更容易操作，將各類感測讀取動作，封裝成方法(Method)定義之。 基於已定義完成之API，為讓開發者更了解ZigBee運作，列出主要七種運作流程， 並說明讀取感測資料之程式控制方式，以及如何擴充新類型感測裝置之方法(Method)，而這些流程與控制方式透過整個網路管理軟體的實作更能體現之。 / ZigBee is a cutting-edge wireless transmission technology, characterized by consuming little power, controlling multiple sensor devices, reacting quickly, and needing low implementing costs, etc. This technology follows IEEE802.15.4 and ZigBee communications protocol. ZigBee-based sensors are widely used in many domains, such as military, natural environments, health, family, and business. Many existing applications of ZigBee use firmware controlling devices, but higher-level application programs are still required to realize some applications. This thesis focuses on how to use Application Programming Interface (API) for controlling devices to collect sensor data and how to solve the problems on reporting data and monitoring all the devices used. In particular, we define the DeviceManager class that controls the set of instructions of device chips and communicates with serial ports as well as determines whether transfer instructions and data by following the standard of ZigBee protocol. For developers' convenience, we further encapsulate different sensor-reading actions into method. Based on the above API definitions, we list seven implementation flows for developers to understand the implementation of ZigBee. This thesis also describes how to use programming control to read sensor data and how to expand new sensor devices. These flows and control methods can be easily achieved by network management software.

本研究使用Android智慧型手機結合無線感測器，提供一個便利學習者攜帶、使用之輔助學習系統，並於智慧型手機提供互動區以利同儕間進行課業討論。系統即時感測、記錄學習者學習行為，並以智慧型手機做為感測資料蒐集及處理中心，為了利用蒐集之學習資料產生最佳推論結果，採用SVM(Support Vector Machine)分類技術進行資料之推論及分析。 本系統之評量及實驗，係藉由讓大學生使用系統，蒐集預測模型建立所需之資料，並了解學習者對本系統之觀感、接受度及系統預測準確率。評量以實驗者前、後測進步情況作為標準，探討與學習情況之關係，藉此了解使用智慧型手機結合感測器輔助學習者進行自我管理，進而提升學習成效。 / We propose a portable, convenient learning assisted system that uses Android Smartphone with wireless sensors. The system senses and collects the data of learning behavior and uses the Smartphone as processing unit to collect statistics. For optimizing inference, we used SVM(Support Vector Machine) technology to analyze the collected data . For evaluating the system , we invited college students to join the experiment and use the system. Therefore, we can collect data to construct the prediction models, knowing better about the accuracy , impression and acceptance of the system , and then realizing the feasibility of using Smartphone with wireless sensors to assist learner promoting learning effect by self-management strategy.