臺灣博碩士論文加值系統

本論文之研究主題是探討卡門濾波演算法在適應性陣列處理之應用。本論
文可分為三部份，在第一部份中 ，我們提出了限制性卡門濾波演算法來
克服傳統非限制性卡門濾波演算法在陣列 監視方向之失真現象。而所謂
限制性卡門濾波演算法是把適應性波束構成之方向限制與微分限制加入卡
門濾波演算法中。第二部份是研討應用卡門濾波演算法於短暫資料之方位
估測。我們發展出具有Thompson之尋找雜訊特徵向量性質之單一模方限制
卡門濾波演算法來估測輸入相關矩陣之全部雜訊特徵向量。另外，建立了
利用卡門濾波演算法之前後向線性預測器，及其方位估測之應用。最後，
本論文探討把均方/預測自迴歸估測器用心臟收縮陣列來實現並應用在方
位估測上。此心臟收縮陣列不僅結構比奇異值分解或特徵分解之心臟收縮
陣列簡化而且當訊號雜訊比高時，效能也和MUSIC演算法相仿。

COVER
1 Introduction
I. I Application of the constrained Kalman algorithm to adaptive beamforming
1.2 Application of the Kalman algorithm to DOA estimation
1.3 Systolic implementation for the DOA estimation
2 Adaptive Beamforming Using the Constrained Kalman Filter
2.1 Introduction
2.2 Constrained Kalman-Type Array Processor
2.2.1 Derivation
2.3 Convergence Analysis
2.4 Simulation Results
2.5 Conclusion
3 Broad-band Kalman-type Adaptive Beamforming Under Directiona and Derivative Constraints
3.1 Introduction
3.2 Biased Constrained Kalman-type Broad-band Adaptive Beamforming
3.3 Unbiased Constrained Kalman-type Broad-band Adaptive Beamforming
3.4 Simulation Results
3.5 Conclusion
4 Kalman-based Spatial-domain Forward-Backward Linear Predictor for DOA Estimation
4.1 Introduction
4.2 Kalman-based Forward-Backward Linear Predictor
4.3 Performance Analysis
4.3.1 Convergence of Weight Vector in the Mean Square
4.3.2 Convergent Behavior of the Mean-Square Prediction Error
4.4 DOA Estimation for Two Coherent Signals
4.5 Simulation Results
4.6 Conclusion
5 Kalman-based Estimator for DOA Estimations
5.1 Introduction
5.2 Kalman-Based Noise-Subspace Estimator
5.2.1 Asymptotic Approach of Weight Vectors Based on Kalman Algorithm
5.2.2 Discussion About the Inflation Method
5.2.3 Comparison of computation complexity
5.3 Simulation results
5.4 Conclusion
6 On the Inflation Method in Adaptive Noise-Subspace Estimator
6.1 Introduction
6.2 An Adaptive Noise-subspace Estimator without Inflation Method
6.3 LMS-type Adaptive Noise-subspace Estimator
6.3.1 Convergence Behavior of the Adaptive Noise-subspace Estimatorwithout inflation method
6.3.2 Discussion about the Inflation Method and Performance Analysis
6.4 Simulation Results
7 A Mlultimode Systolic Array Implementation for DOA Esl
7.1 Introduction
7.2 A Smoothing/Prediction Autoregressive (S/P AR) Estimator
7.3 Multimode systolic array implementation
7.4 Fast Givens algorithm for systolic array implementation . . .
7.5 Simulation results
7.6 Conclusion
8 Conclusions
Appendices
A Convergence of the Biased Constrained Kalman Algorithm
B Iteration Equations of the Unbiased Constrained Kalman Algorithm
C Convergence of the Unbiased Constrained Kalman Algorithm
D Convergence of the Kalman-based Noise-Eigenvector Estimator
bliography