在傳感器數(shù)量有限且傳感器分布空間不足等有限資源的情況下,如何尋找比傳感器更多的源(稱為欠定DOA估計)非常重要。本論文的主要重點是在陣列處理中考慮欠定的觀測模型,其中陣列接收的信號源的數(shù)量可以大于物理陣元的數(shù)量。顯然,均勻線性陣列(ULA)中存在更高程度的冗余,應(yīng)該將其解決為最小冗余。用于信號接收和空間譜估計的陣列設(shè)計中反復(fù)出現(xiàn)的問題是如何有益地部署稀疏陣列的陣元來最佳的采樣空間頻譜。因此,在確保最佳空間樣本采樣間隔的同時,如何尋找具有連續(xù)整數(shù)和最小冗余的陣元位置是一個挑戰(zhàn)。本研究的重點在于合理的部署陣列,來檢測比陣元數(shù)目更多的信號源,利用陣列結(jié)構(gòu)來獲得盡可能高的分辨率并增加可檢測的信號源數(shù)量。為了使陣列配置具有成本效益,對于給定的成本,最大化陣列孔徑尺寸同時放置最少傳感器來實現(xiàn)更高分辨率是先決條件。本論文的主要貢獻(xiàn)是從有限數(shù)量的陣元來增加陣列孔徑長度,以達(dá)到盡可能高的分辨率,減少相關(guān)矩陣中的冗余元素。最小冗余陣列(MRA)能夠?qū)崿F(xiàn)最小冗余,但是這類陣列需要窮舉搜索程序來確定陣元的位置。此外,沒有系統(tǒng)的方法或指定的公式來設(shè)計具有未知陣元分布的MRA。共陣在欠定DOA估計的稀疏結(jié)構(gòu)中起著關(guān)...

【文章頁數(shù)】：228 頁

【學(xué)位級別】：博士

【文章目錄】：
摘要
ABSTRACT
Notation
Chapter 1 Introduction
    1.1 Background
    1.2 Literature Review
    1.3 Main Task
    1.4 Outline and Scope of this Thesis
Chapter 2 Array Model and Related Materials
    2.1 Background
    2.2 The Array Signal Model
    2.3 Autocorrelation and Spectral Estimation
    2.4 Parametric Estimation
    2.5 Spectral Estimation
    2.6 Sampling and Underdetermined Estimation
    2.7 Different Array Model
        2.7.1 Linear Arrays
        2.7.2 Planar Arrays
        2.7.3 Circular Arrays
        2.7.4 Coprime Arrays
    2.8 Different fields of view and frequency bands
        2.8.1 Far Field
        2.8.2 Near Field
        2.8.3 Narrow Band
        2.8.4 Wide Band
    2.9 Chapter Summary
Chapter 3 Popular Methods for DOA Estimation
    3.1 Background
    3.2 Classical Beamforming Method
    3.3 MUSIC Algorithm
    3.4 ESPRIT Algorithm
    3.5 Genetic Algorithm
    3.6 Particle swarm optimization (PSO) algorithm
    3.7 Sparse Bayesian Learning Method
    3.8 Chapter Summary
Chapter 4 Virtual Extension exploiting Difference and Sum
    4.1 Introduction
    4.2 Signal Model
    4.3 Proposed Methodology
        4.3.1 Extension of Vitual Arrays Exploiting Difference and Sum Co-array
        4.3.2 MUSIC based DOA estimation
    4.4 Simulation Results
        4.4.1 Case Ⅰ: RMSE for different SNR
        4.4.2 Case Ⅱ: RMSE for different number of Snapshots
        4.4.3 Case Ⅲ: RMSE for different number of Sources
    4.5 Chapter Summary
Chapter 5 Novel Array Structure using Translocation，Axes Rotation and Compression
    5.1 Introduction
    5.2 Signal Model
    5.3 Proposed Methodology
        5.3.1 Conventional Coprime Array Configuration
        5.3.2 Proposed Coprime Array Configuration
        5.3.3 The Difference Co-array of Proposed Method
        5.3.4 Interpolation with Iterative Power Factorization
        5.3.5 MUSIC based DOA estimation
    5.4 Simulation Results
        5.4.1 Case Ⅰ: RMSE for different SNR
        5.4.2 Case Ⅱ: RMSE for different number of Snapshots
        5.4.3 Case Ⅲ: RMSE for different number of Sources
        5.4.4 Number of Lags vs Number of Sensors
    5.5 Chapter Summary
Chapter 6 Novel Array Structure unifying Trio Subarray and FOD
    6.1 Introduction
    6.2 Signal Model
    6.3 Proposed Methodology
        6.3.1 Conventional Coprime Array Configuration
        6.3.2 Proposed Coprime Array Configuration
        6.3.3 The Fourth Order Difference Co-Array of Proposed Method
        6.3.4 Sparse Baysian Learning Based DOA Estimation
    6.4 Simulation Results
        6.4.1 Case Ⅰ: RSME for Different SNR
        6.4.2 Case II: RMSE for Different Number Of Snapshots
        6.4.3 Case Ⅲ: RMSE for Different Number of Sources
        6.4.4 Case Ⅳ: RMSE for Lower Angular DOA Estimation
        6.4.5 Number of Lags vs Number of Sensors
    6.5 Chapter Summary
Chapter 7 Novel Array Structure comprising Triplet Coprime Array
    7.1 Introduction
    7.2 Signal Model
    7.3 Proposed Methodology
        7.3.1 Proposed Coprime Array Configuration
        7.3.2 The Second Order Difference Co-Array of Proposed Method
        7.3.3 MUSIC DOA Estimation
        7.3.4 Sparse Baysian Learning Based DOA Estimation
        7.3.5 The Fourth Order Difference Co-Array of Proposed Method
        7.3.6 Sparse Baysian Learning Based DOA Estimation
    7.4 Simulation Results
        7.4.1 Case Ⅰ: RSME for Different SNR
        7.4.2 Case Ⅱ: RMSE for Different Number Of Snapshots
        7.4.3 Case Ⅲ: RMSE for Different Number of Sources
        7.4.4 Case Ⅳ: RMSE for Lower Angular DOA Estimation
        7.4.5 Number of Lags vs Number of Sensors
    7.5 Chapter Summary
Chapter 8 Conclusion
    8.1 Summary
        8.1.1 VECADS
        8.1.2 CATARCS
        8.1.3 VEFODCI
        8.1.4 TiCADD
    8.2 Future Works
Bibliography
Appendix A
Acknowledgements
Publications



本文編號：4031989