本論文於無線感測網路基礎下設計一情境感知服務(Context-Aware Service)系統並實作應用於居家環境之裝置控制功能。此系統之無線感測網路以ZigBee無線模組與多個不同感測器(Sensor)所建構而成，藉由各感測器的功能屬性蒐集環境變數資訊，並使用ZigBee無線模組為其傳輸媒介，將資訊傳送至情境感知服務系統；經情境管理者智慧判斷與推論後，提供服務給使用者。
情境感知系統應具有瞭解使用者意圖之能力。除了藉由感測器之協助收集使用者周遭環境資訊外，情境管理者(Context manager)亦內建規則，當規則成立時，便提供服務。但這種作法仍因使用者習性不同或其他因素變化而充滿了不確定性，所以本研究另基於貝氏網路(Bayesian network)理論基礎下建立一推論機制。此推論機制之模組，藉由使用者回饋之歷史經驗作為參考依據，將影響條件之變數區分為提供服務與不提供服務等類型，計算使用者所處環境資訊與時間資訊之條件機率，用以評估是否適合提供服務。
本系統除設計情境管理者的規則(rules)判斷與推論機制之預測，以提高使用者真正所需服務之正確度外，並利用實作之成果進行測試。由測試結果得知，本論文設計之情境感知服務系統能判斷使用者在感測範圍內的服務需求，推論出有利於使用者的決定並提供服務，讓使用者不需要自行動手操作便能使用日常生活周遭不同裝置提供的功能。

This thesis presents the design of a context–aware service system based on wire- less sensor network and the implementation of home services for some described app- lication scenarios. The wireless sensor network employed in this design is ZigBee which consists of coordinator, sensor nodes and device nodes. Sensor nodes collect environmental information and transmit the collected information to the context aware service system through coordinators. Based on the received context information and some pre-set rules, the context aware service system infers and determines the suit- able service which might be desirable for the user. The device nodes control devices according to the commands issued from the service system.
A context aware service system should have the ability to understand user’s needs. In addition to collecting user’s surrounding environment information with the assistance of the sensor devices, some context managers build the rules to provide ser- vice when tenable in ruling. However, this way of providing services is still full of uncertainty. To address this issue, we propose a prediction mechanism based on Bayesian network's theory. The prediction model divides influence condition parame- ter into two parts: providing service and not providing service. Using user feedback in the previous experience as basis of consulting, the prediction model then calculate the probability of the service desirable by user’s living in this time zone and in the envi- ronmental area to assess whether is suitable for providing service or not.
The implementation in this thesis includes a basic context-aware service system, the ZigBee wireless sensor network, and three example home services, namely tempe- rature service, TV service and bedroom service. From the result of testing, it can be seen that the sensors work well and the inference obtained from the prediction mecha- nism in most cases can provide services desired by the user. The design of this context-aware service system therefore is effective.

[1] Schmidt Albrecht, et al., “Advanced interaction in context,” HUC99, LNCS 1707, Springer Verlag.
[2] Dey Anind , “Understanding and using context,” Personal and Ubiquitous Computing Journal, pp. 5(1): 4-7, 2001.
[3] Di Zheng, Jun Wang, Weihong Han, Yan Jia, Peng Zou, “Towards A Context-Aware Middleware for Deploying Component–Based Applications in Pervasive Computing,” Conference on Grid and Cooperative Computing,
pp. 454-457, 2006.
[4] Qin Huaifeng, Zhou Xingshe, “Context Aware Sensornet,” First International Conference on Semantics, Knowledge and Grid, ACM, pp. 1-7, 2005.
[5] Kaori Fujinami, Tetsuo Yamabe, Tatsuo Nakajima, “Take me with you ! A Case Study of Context-Aware Application integrating Cyber and Physical Spaces,” ACM Symposium on Applied Computing, pp. 1607-1614, 2004.
[6] Renato Jorge Caleira Nunes, Jose C. M. Delgado, “An Internet Application for Home Automation,” Mediterranean Electrotechnical Conference, MEleCon ,
pp. 298-301, 2000.
[7] Guanling Chen, David Kotz, “A Survey of Context-Aware Mobile Computing Research,” Technical Report TR 2000-381, Darmouth College.
[8] Sven Meyer, Andry Rakotonirainy, “A Survey of Research on Context-Aware Homes,” Conferences in Research and Practice in Information Technology, Vol 21, pp. 159-168, 2003.
[9] David Heckerman, “A tutorial on learning with Bayesian networks,” Technical Report MSR-TR-95-06, Microsoft Research, Microsoft Corporation,
pp. 301-354 ,1995.
[10] Judea Pearl, “probabilistic reasoning in intelligent systems: Networks of plausible Inference,” Morgan Kaufmann Publishers, San Francisco, 1988.
[11] Sang-Hak Lee, Tae-Choong Chung, “System Architecture for Context-Aware Home Application,” IEEE WSTFEUS, pp. 149-153, 2004.
[12] JungRae Kim, JaeDoo Huh, “Context-AWare Services Platform Supporting Mobile Agents For Ubiquitous Home Network,” IEEE International Conference Advanced Communication Technology, pp. 136-139, 2006.
[13] HyungJik Lee, JiEun Park, EunJung Ko, JeunWoo Lee, “An Agent-based Context-Aware System on Handheld Computer,” IEEE International Conference on Consumer Electronics, pp. 229-230, 2006.
[14] Tao Gu, Hung Keng Pung, Da Qing Zhang, “A MiddleWare for Build Context-Aware Mobile Services,” IEEE Vehicular Technology Conference,
pp. 2656-2660, 2004.
[15] Jonghwa Choi, Dongkyoo Shin, Dongil Shin, “Research and Implementation of the Context-Aware Middleware for Controlling Home Appliances, ” IEEE International Conference on Consumer Electronics, pp. 301-306, 2005.
[16] Hyunjeong Lee, Jongwon Kim, Jaedoo Huh, “Context-Aware based Mobile Service for Ubiquitous Home,” IEEE International Conference Advanced Communication Technology, pp. 1851-1854, 2006.
[17] Mark Weiser, “Hot Topics: Ubiquitous Computing” IEEE Computer, pp. 71-72, October 1993.
[18] Neil Martin, Fenton Norman, Tailor Manesh, “Using Bayesian Networks to model Expected and Unexpected Operational Losses,” Risk Analysis: An International Journal, pp. 963-972, 2005.
[19] Bill N. Schilit, Norman Adams, Roy Want, “Context-Aware Computing Application,” IEEE Workshop on Mobile Computing Systems and Application, pp. 1-7, December 1994.
[20] Microchip Technology Inc, “Microchip stack for the ZigBee protocol,” 2006,
http://www.microchip.com/.
[21] Zigbee alliance, http://www.zigbee.org/.
[22] MySQL Reference Manual, http://dev.mysql.com/doc/refman/5.1/en/index.html.
[23] 傅立成, “智慧型家庭之動態個人偏好學習系統及環境感知服務提供,” 碩士論文, 國立台灣大學, June, 2006.