至今以電池為動力來源的無線感測器所構成的無線感測器網路已經被廣泛應用在多樣化的環境偵測上,由於無線感測器在傳輸訊號時會消耗電池的電力,所以電池的壽命長短限制了感測器的工作時間,因此如何降低無線感測器因傳輸所消耗的電力是一項值得探討的議題。本研究在此設計出高能源效率的演算法來將降低錯誤率以及能源消耗。 在傳統上我們用1bit來記錄所觀測的訊號,也就是二維的作法,本論文使用多維訊源編碼機制應用於無線感測器網路中,將感測器所觀測到的訊號進行編碼分區的動作,將二維拓展為多維,再透過決策融合中心的軟性決策與硬性決策兩種機制來降低其錯誤率以及節省能源消耗。在傳統的無線感測器網路上只作二維的傳輸方式,因此本論文提出利用多維的傳輸方式,再根據融合中心的兩種決策演算法來降低傳輸錯誤率,而經實驗與理論證實此兩機制確實能達到比傳統上有著更低的傳輸錯誤率也能達到節省能源的目標。
The wireless sensor network(WSN) structured by the battery-powered sensor has been widely used in detecting a variety of environments. The sensor consumes energy while transmitting signals and hence the life of battery directly limits the working hours of the sensor. Therefore, reducing the power consumption of wireless transmission is an issue worth exploring. This study focuses on the design of energy-efficient algorithm to reduce energy consumption while the goal of better error performance is also achieved. In the conventional WSN systems, 1 bit is used to record the observation of the environment. In this thesis, we use M-ary source coding scheme to divide and to encode the observation to expand the binary algorithm to an M-ary algorithm. Soft-decision and hard-decision rules are both adopted at the fusion center. The results show that lower error rates and energy consumption are achieved via the proposed M-ary source coding scheme.