在機載雷達的使用場景中，地雜波和干擾器的位置與速度分布會隨著時間快速變化，這些干擾訊號和雜訊在接受訊號中會大大地影響我們對目標訊號檢測成功率。在次世代的機載雷達上，相控陣列雷達具備多通道多功能、低反應時間、參數可適配多場景及可靠度高等特性，其中多通道多功能的特性也使得諸如Adaptive Beamforming (ABF) 和Space-Time Adaptive Processing (STAP) 等技術能夠使用以消除機載雷達使用場景的干擾與雜訊。
本篇論文將對時空雙域自適應訊號處理 (STAP) 進行研究，從脈衝都卜勒相控陣列雷達的Data-Cube資料集特性開始說明，開發多種STAP降複雜度的核心演算法，包括EFA STAP、JDL STAP和Σ-∆ STAP；並且針對對空移動目標和對地移動目標的場景效能指標進行模擬，最後模擬並評估出實時運算的可能性。模擬結果顯示，STAP演算法確實可以有效地將地雜波與干擾器的訊號消除，並且提升目標所在的能量；降複雜度的STAP也能夠在系統效能不損失太多的前提下，符合實時運算複雜度的標準。
此外，本文也會對STAP系統應用的架構進行優化，完整說明STAP產出的權重如何應用在Angle-Doppler Map以及Range-Doppler Map上，以利後續待開發的Constant False Alarm Rate (CFAR) Detection作使用。

It is common to refer to multiple correlated sources of noise as interference and use the signal-to-interference-plus-noise ratio (SINR) as one of key factors influencing detection performance. Airborne radars must possess the capability to suppress both clutter and jamming, which are two major sources of unwanted signal, to near or below the noise level. Pulsed Doppler radars with an active electronically scanned array (AESA) can be designed to provide signal diversity in the form of the elements of the array antenna and multiple pulses in a coherent waveform interval, facilitating a number of functions, including: adaptive beamforming (ABF) to counter jamming and space-time adaptive processing (STAP) to remove clutter and to detect and track air or ground moving targets.
STAP involves dynamically adjusting a two-dimensional space-time filter response in an attempt at maximizing the output SINR, offering the potential to improve detection of the slow-moving target, and detection in combined clutter and jamming environments. This paper analyzes three type of reduced-dimension STAP approaches: EFA (Extended Factor Approach) STAP, JDL (Joint Domain Localized) STAP and Σ-∆ (Sigma-Delta) STAP, starting from a common set of assumptions and system parameters, and data-cube simulation in homogenous environments. To this end, simulation results are presented to illustrate various performance metrics with definitions and formulations, including angle-Doppler map, range-Doppler map, improvement factor, and SINR loss.
In addition, the structures of the algorithms above will be adjusted to reduce the complexity of the signal processing compared to that have been described in the reference. Computing time of each algorithm on the basis of real-time computing standard and the performance in air moving-target indication or ground moving-target indication scenario will also be evaluated.

摘要 1
Abstract 2
目錄 4
圖目錄 7
表目錄 10
第1章 緒論 1
1.1 前言 1
1.2 研究動機 2
1.3 相關研究 3
1.4 論文成就與貢獻 4
1.4.1 STAP系統架構的逆向工程 4
1.4.2 STAP核心演算法複雜度的優化 4
1.4.3 STAP 核心演算法在不同場景的效能評估 4
第2章 系統模型 6
2.1 座標系統 6
2.2 陣列因子 9
2.3 波束成型 10
2.4 Data-Cube資料集 13
2.4.1 目標訊號 16
2.4.2 熱雜訊訊號 19
2.4.3 地雜波訊號 20
2.4.4 敵機干擾訊號 24
第3章 系統架構 26
3.1 STAP演算法 26
3.2 降複雜度STAP演算法 29
3.3 EFA STAP 31
3.3.1 Subarray Partition 33
3.3.2 Training Strategy 33
3.3.3 Dimension Reduction 34
3.3.4 Covariance Matrix Estimate 35
3.4 Weight Application 37
3.4.1 Angle-Doppler Response 37
3.4.2 Range-Doppler Response 38
3.4.3 Weight Normalization 39
3.4.4 Sliding Window Training 41
3.4.5 Block-Wise Training 42
3.5 JDL STAP 45
3.5.1 Subarray JDL STAP 47
3.5.2 Full Array JDL STAP 49
3.6 Σ-∆ STAP 50
3.6.1 Subarray Partition for Σ-∆ STAP 51
3.6.2 Sigma and Delta Beamforming 52
第4章 模擬結果 56
4.1 檢測場景參數設定 56
4.2 演算法效能指標 59
4.2.1 Improvement Factor 59
4.2.2 SINR Loss 59
4.3 演算法效能評估 61
4.3.1 演算法效能驗證 61
4.3.1 Airborne Moving Target Indication 67
4.3.2 Ground Moving Target Indication 68
4.4 演算法實時運算評估 71
第5章 結論 74
參考文獻 75

[1] S. Moore, “Review of The State of The Art of UK AESA Technology and The Future Challenges Faced,” 2008.
[2] A. D. Brown, Electronically Scanned Arrays MATLAB Modeling and Simulation, 2012.
[3] B. R. Mahafza, Radar Signal Analysis and Processing Using Matlab, 2009.
[4] J. Ward, “Space-Time Adaptive Processing for Airborne Radar,” 1994.
[5] J. A. S. W. A. H. Mark A. Richards, Principles of Modern Radar: Vol. 1 Basic Principles, 2010.
[6] T. B. Hale, Airborne Radar Interference Suppression Using Adaptive Three-Dimesion, 2002.
[7] W. B. Diego Cristallini, “Strategies for Sub-optimal Air to Air STAP in Forward Looking Configuration,” 2010.
[8] J. A. S. William L. Melvin, Principles of Modern Radar: Vol. 2 Advanced techniques, 2014.
[9] W. Bürger, “Space-Time Adaptive Processing: Algorithms,” 2006.
[10] W. L. M. B. H. Audrey S. Paulus, “Performance and Computational Trades for RD-STAP Algorithms in Challenging Detection Environments,” 2016.
[11] L. C. Hong Wang, “On Adaptive Spatial-Temporal Processing for Airborne Surveillance Radar Systems,” 1994.
[12] M. C. W. Raviraj S. Adve, “Joint Domain Localized Processing Using Measured Spatial Steering Vectors,” 1998.
[13] D. K. B. Samuel M. Sherman, Monopulse Principles and Techniques, Second Edition, 2011.
[14] R. A. S. M. C. W. H. W. Y. Z. Russell D. Brown, “STAP for Clutter Suppression with Sum and Difference Beams,” 2000.
[15] J. A. S. William L. Melvin, Principles of Modern Radar: Vol. 3 Radar Applications, 2013.
[16] Y. Z. H. W. John Maher, “Space-time Adaptive Processing with Sum and Multiple Difference Beams for Airborne Radars,” 1999.