- 學位論文
用於智慧天線型感知無線電之三維空-時-頻感測器設計與實現
Design and Implementation of Three-Dimension Space-Time-Frequency Sensing for Cognitive Radios with Smart Antennas
摘要
感知無線電的目的有二，其一，感測頻域中的閒置頻譜，其二，對閒置頻譜再利用以提昇頻譜使用效率。然而，傳統上只配置單根天線的頻譜感測器因缺少空間資源，因此無法鑑別接收訊號的入射方向，以再提昇感知容量。另外，單天線頻譜感測器經常無法偵測出在強干擾訊號下低功率的主要使用者(Primary User, PU)，大幅造成感知系統的不可靠性。為解決上述問題，Gardner提出結合多天線接收與頻域感測技術，稱為SCORE演算法[52]，不過SCORE演算法運算複雜度高，難以硬體實現，為降低複雜度，Dou提出適性Adaptive SCORE演算法[2]，以利於實現。最近，為使偵測效能再提昇，時域的資訊也引入，Chopra 提出限制型雙邊可適性最小均方 (constrained doubly adaptive LMS，C2LMS)演算法[53]。C2LMS是用於智慧天線型感知無線電之低複雜度三維空-時-頻感測器演算法。本論文基於C2LMS的架構，並將其擴展至寬頻感知系統上，提出寬頻C2LMS (Wideband C2LMS, WB-C2LMS) 感測演算法。模擬結果顯示，本論文提出之感測方法可使寬頻感知系統下弱PU在強干擾訊號的影響下仍能被偵測出。此外，本論文也採用Xilinx的System Generator實現WB-C2LMS演算法，以驗證本論文提出的演算法適用於硬體電路的實現。
並列摘要
The purpose of the cognitive radio has two features: the first feature is to detect the white space of the spectrum, and the second feature is to re-use the white space of the spectrum for improving the spectrum efficiency. However, the conventional spectrum sensing with single antenna is lack of the spatial resource for distinguishing the arrival angle of the received signal, and thus the cognitive capacity cannot further be improved. Besides, the spectrum sensor with single antenna cannot detect the weak-power primary user (PU) under the scenario of the strong interference, and thus the system reliability is reduced significantly. To solve the above problem, Gardner proposed the spectrum sensing technique combined with the multiple-antenna reception, called the Spectral Coherence Restoral (SCORE) algorithm [52]. However, the computational complexity of the SCORE algorithm is too high to be implemented. Hence, Dou proposed the adaptive SCORE algorithm for hardware implementation [2]. Recently, Chopra proposed constrained doubly adaptive LMS (C2LMS) algorithm by introducing the temporary information to improve the detection rate [53].The C2LMS is the low-complexity three-dimensional space-time-frequency (3D-STF) sensor algorithm with smart antennas for the cognitive radio system. This thesis is based on the C2LMS structure, and then we present the broadband C2LMS algorithm (Wideband C2LMS, WB-C2LMS) for the broadband cognitive radio system. Simulation results demonstrate that the proposed 3D-STF spectrum sensing algorithm can favor the weak-power PU under the scenario of the strong-power interference. Moreover, this thesis also exploits the System Generator of Xilinx to implement the WB-C2LMS algorithm to verify the hardware feasibility.