本論文為針對數位電視信號研究頻譜偵測技術,目的為提昇偵測技術的靈敏度,適用於低訊雜比之環境。本論文中有詳細的介紹習知頻譜感測技術,例如能量偵測(EG)、協方差絕對值偵測(CAV)、特徵值偵測(MME)、循環穩態偵測(CS)、匹配濾波器偵測技術(MF)等不同偵測方法,並深入分析其優缺點。本論文亦對CAV偵測技術提出改進架構,使得比原傳統的CAV複雜度較低並且性能也較佳;並對MF偵測做改良,不因外在環境變化而能夠確保偵測性能穩定之需求。為了更進一步提昇性能,本論文亦提出兩種改進技術,其一為具決策與轉傳之中繼器合作式偵測頻譜技術;其二為具放大轉傳之中繼器合作式偵測頻譜技術,兩種技術在感知中繼器個數越多時,其效能也越佳。經由電腦模擬驗證得知所提出之方法具有明顯的性能提昇,可適用於低訊雜比之訊號偵測,並且該結果吻合所推導之合作式偵測法則理論性能,得以證實本論文所提出方法之正確性。接續,利用軟體無線電平台(SDR)進行實際測試,將數位電視訊號實際發射並進行訊號偵測,完成使用不同偵測法實現於硬體平台之驗證。最後,經由上述SDR平台驗證測試,可證實本論文所提出之技術可行性佳及具有實用價值。
The purpose of this thesis is to research the spectrum sensing technology for digital TV signal, and enhance the sensitivity of the detection of low signal-to-noise ratio (SNR) signal. In this thesis, the traditional spectrum sensing technology has been introduced in detail with advantage and disadvantage analysis, e.g., energy detection (EG), covariance absolute value (CAV) detection, maximum minimum eigenvalue (MME), cyclostationary detection (CS), and matched filter detector. This thesis also proposes the improvement architecture in CAV detection, which provides low complexity advantage and better performance than the traditional CAV detector. Next, the adaptive MF detection scheme is proposed to ensure the stability of the detection performance over the time-varying environment. Moreover, in order to enhance detection performance, two improved technologies are proposed, i.e., decision-forward and amplify-forward cooperative detection technologies. Note that the performances are better as the number of cognitive repeater increases. Simulation results confirm that the proposed methods can enhance performance obviously, which can be used for the detection of low SNR signal. Furthermore, the simulated results can approach the theoretical performance of the cooperative detection scheme, which can prove the proposed method with correctness. Finally, using SDR platform for verification, the digital TV signal can be detected by the different spectrum sensing methods. It is confirmed that the proposed detection technique is available and realizable in realistic cognitive radio environments.