由於無線感測網路的進步，使得物體與物體之間透過網際網路相互連接，成為物聯網（Internet of Things, IOT），大量運用在定位及環境監控等。本研究目的在提出一套利用少數網路攝影機和無線感測器結合的監控系統(Monitor and Control System, MCS)，並輔以人工智慧技術。使用模糊控制系統，改善現有定位方式，提升定位準確度。感測裝置部分，透過PSoC嵌入式基板，連結溫、濕度感測晶片並運用Uart方式將訊號轉換給ZigBee網路模組，傳送至後端伺服器WiNOC系統接收。針對異常狀況及節點移動的狀況，啟動視訊攝影機進行監控。本研究採用MCS監控系統與傳統監視系統進行比較，結果顯示本研究具有優秀之監控效果。
本研究並以Network Simulater2(NS2)進行實驗的模擬，在感測系統部分，探究ZigBee網路及其安全狀況並進行分析，進行網路攻擊模擬，觀察遭受攻擊對正常的服務影響，並彙整網路資訊。
最後透過人工智慧的類神經網路系統，透過不同的隱藏層神經元節點與不同的訓練次數，找尋最好的類神經網路模型，做為網路攻擊入侵偵測系統。比較各種不同的佇列機制及功用，透過NS2網路模擬的方式，模擬在遭受攻擊時的正常封包遺失量，探究其節點防禦能力。
According to the advances in wireless sensor networks, making between objects and objects connected to each other through the Internet (Internet of Things, IOT). It extensive uses in positioning and environmental monitoring. The purpose of this study is a network using a small number of cameras and sensors combined with wireless monitor and control system (MCS), and supplemented by artificial intelligence technology. It uses fuzzy control systems improve the positioning accuracy. Through the PSoC embedded system, we link and use sensor chip detect the temperature and the humidity. The signal conversion to the ZigBee network by UART, and sent to the server, WiNOC system. In the abnormal condition and status of the moving nodes, it starts the video camera and monitor. In this study, we compare the MCS and the traditional surveillance systems. The MCS has shown excellent results of monitoring.
In this study, we use Network simulater 2(NS2) to simulation. In the part of the sensor network, we explore the ZigBee network and analyze the network situation when it under attack. We observe how the offenses affect to the normal service and collecte the data.
Finally, we use the artificial intelligence nural network system. Through the different nodes in the hidden neurons layer and the different training times, to find the best neural network model in Intrusion Detection System, IDS. And we compare the function of the different queues. We simulate by NS2 to find the amount of the normal packets loss when the node is under attack.
