論文提出一種於瑞雷衰減通道(Rayleigh Fading Channel)下,採取分散決策的感知無線電 (Cognitive Radio Network)的模擬信號頻譜偵測模型架構。本文透過開關鍵控(On-off Keying)訊號模擬信號接收偵測的情況,並且在感測器端使用了一種同調振幅解調的偵測技術,聯合接收中心端採取多數決策合成的決策方案來統合信號。最後與傳統的分集合成技術:選擇性合成(Selection Combining)做為比較,證明在偵測信號震幅並且限制誤警機率(Probability of False-alarm)為一定數的狀況下,本文所提出的分散決策可以提供比選擇性合成更好的遺漏機率效能。
In this thesis, a signal spectrum sensing scheme for Cognitive Radio Network based on distributed detection over Rayleigh fading channel is provided. On-off keying signal modulation is simulated to signal detection behavior, a sensing technique which is called Coherent Amplitude Detection is used in sensor node, and Majority-Voting Combining is proposed for fusion center in this thesis. Finally, it is showed that the Miss-Detection probability for Majority-Voting Combining based on distributed detection is better than Selection Combining if the False-Alarm probability is limited to a constant.