目前常見的無線感測網路，多採用單一維度且無即時串流型態的感測器，如 溫度、溼度或是煙霧等。聲音感測則呈現高度即時性的資料串流特性，當其在無 線網路傳輸時通常需要低延遲的傳輸性能。本研究即針對聲音偵測、訊號處理以 及無線聲音傳輸的應用，設計一個低功耗的無線聲音感測網路。 我們參考IEEE 802.15.4 與Zigbee 協定，針對無線聲音傳輸特性，設計一個 無線聲音感測網路MAC 層協定，以提供精簡且高效率的通訊；我們規畫了一個 無線聲音傳輸的封包結構，並提出新的無線聲音同步方法，以改善無線訊號傳輸 過程封包遺失所造成的聲音不連續性。最後我們設計一組電源管理策略，可以有 效達到節能的目標。 我們在ARM Cortex M3 嵌入式微處理器平台上實作了此一無線聲音感測網 路並進行系統整合驗證。實驗結果顯示，我們所設計的系統對於聲音的延遲與同 步特性有著優異的特性；相對於傳統的無線資料傳輸，我們所使用的封包精簡約 50%，因此可以減少傳輸次數，提高效能，同時也節省耗電量。At present, common wireless sensing network(WSN), most are based on a single dimension and non real-time streaming types of sensors, such as temperature, humidity, or Smoke etc. Sound sensing presents the highly instantaneity of real time data streaming characteristic. When it in wireless networks usually require low latency transmission performance. This research focuses on sound detection, signal processing and wireless sound transmission of the application, and design a low power wireless sound sensor network We refer to the IEEE 802.15.4 and Zigbee protocol, features for wireless sound transmission, and design a wireless sound sensor network MAC layer protocol to provide a streamlined and high efficient communication. We have planned a wireless sound transmission packet structure, and propose a new s wireless sound synchronization method, to improve the wireless signal transmission voice packet loss caused by the discontinuity. Finally, we design a power management strategy can effectively achieve the energy conservation. We implemented wireless sound sensor network and system integration verification with the ARM Cortex M3 embedded microprocessor platform. Experimental result shows, We designed the system for the sound delay and synchronization with superiority features, Compared with traditional wireless data transmission, we reduced about 50% packet size, therefore transmission times can be reduced, enhances the efficiency and saving the power consumption.
