隨著5G通訊技術邁向毫米波(mmWave)頻段，如何在頻譜資源稀缺且干擾複雜的環境中維持高品質傳輸，已成為關鍵技術挑戰。本論文研製了一套28GHz毫米波混合主動陣列模組(Hybrid Active Array Module)，並以此硬體平台為基礎，深入探討基於空間零點(Spatial Nulling)與MIMO (Multiple-Input Multiple-Output)處理之干擾抑制技術。 本研究首先針對陣列天線在多訊號環境下的物理特性進行量化分析。實驗顯示，雖然雙極化配置能提供部分隔離，但在缺乏主動干擾抑制的情況下，殘餘的跨極化干擾仍導致EVM (Error Vector Magnitude)劣化，無法滿足高階調變需求。為突破此物理限制，本論文提出了一套結合MUSIC (Multiple Signal Classification)到達角估測、MMSE (Minimum Mean Square Error)波束成形與MIMO運算的的整合式訊號處理架構。 在實測驗證方面，利用2 port 1x8雙通道子陣列架構進行了兩大關鍵場景評估，針對低軌衛星長距離傳輸，驗證了演算法在功率受限環境下的增益維持與訊號重建能力；針對地面基地台多用戶場景，則展現了演算法在干擾受限環境下，形成深度空間零點以極大化訊號干擾(SIR Signal-to-Interference Ratio)的優勢。綜合而言，本論文成功實現了一套軟硬體整合的毫米波驗證平台，證實了透過MMSE建立高訊雜比鏈路與MIMO優化解調精度的協同運作，能有效克服同頻干擾與路徑損耗，為未來高容量、高可靠度的毫米波通訊系統提供了具體的工程實踐與設計參考。 ;As 5G communication technology advances towards the millimeter-wave (mmWave) spectrum, maintaining high-quality transmission in spectrum-scarce and interference-complex environments has become a critical technical challenge. This thesis presents the development of a 28GHz mmWave Hybrid Active Array Module. Based on this hardware platform, the study investigates and verifies interference suppression techniques based on Spatial Nulling and MIMO (Multiple-Input Multiple-Output) processing. The study first quantifies the interference characteristics of antenna arrays in multi-signal environments. Experimental results show that while dual-polarization configurations provide partial isolation, residual cross-polarization interference still causes significant degradation in EVM (Error Vector Magnitude) without active interference suppression, failing to meet the requirements for high-order modulation. To overcome this physical limitation, this thesis proposes an integrated signal processing architecture combining MUSIC (Multiple Signal Classification) for Direction of Arrival (DOA) estimation, MMSE (Minimum Mean Square Error) beamforming, and MIMO algorithms. In terms of experimental verification, a 2-port 1x8 dual-channel subarray architecture was employed to evaluate performance in two key scenarios. For Long-Distance LEO Satellite Transmission, the results verified the algorithm′s capability to maintain gain and reconstruct signals under power-limited conditions. For the Ground Station Multi-User Scenario, the study demonstrated the algorithm′s advantage in forming deep spatial nulls to maximize the Signal-to-Interference Ratio (SIR) within interference-limited environments. In conclusion, this thesis successfully implements a software-hardware integrated mmWave verification platform. It confirms that the synergistic operation of establishing high-SNR links via MMSE and optimizing demodulation precision via MIMO can effectively overcome co-channel interference and path loss, providing concrete engineering practices and design references for future high-capacity, high-reliability mmWave communication systems.
