在恆定模量約束下使用動態選點方法達到低峰值旁辦位準的均勻矩形陣列波束圖型合成

Abstract

波束成形在衛星通信中至關重要，因為它在克服長距離傳輸所造成的巨大路徑損耗方面發揮了關鍵作用。為了在廣泛區域內同時為用戶終端提供服務並提高下行鏈路容量，衛星廣播應用中需要設計具有寬波束的波束成形器。另外，對於最小化發射波束的功率洩漏，從而防止對非目標用戶的干擾，在波束圖型合成中實現低峰值旁瓣位準也是至關重要。在衛星波束成形器的設計中，必須考慮恆定模量約束，以使功率放大器在接近飽和點的同時保持在線性區域運作，從而實現最大效率。本篇論文研究了均勻矩形陣列發射波束成形器設計問題，其目的是在滿足恆定模量約束的前提下，最小化發射波束圖型的低峰值旁瓣位準。本篇論文提出了稱為動態選點的新方法，通過找到局部峰值進行壓抑，以確保在連續空間域內抑制低峰值旁瓣位準。使用此方法可以減少約束的數量，縮短計算的時間，並獲得更低的低峰值旁瓣位準。針對非凸函數約束的問題，最佳化問題被重新制定為帶有秩等於一約束的半正定規劃，並通過凸函數迭代算法求解。模擬結果顯示了對比其他已存在方法，此方法在滿足恆定模量約束之下抑制低峰值旁瓣位準方面的優勢。

Beamforming is essential in satellite communication (SatComs) since it plays a crucial role in overcoming great path loss caused by long-distance transmission. To achieve higher downlink capacity while simultaneously serving user terminals over wide areas, broadened beam beamformer design is desired in satellite (SAT) broadcast applications. Furthermore, achieving a low peak sidelobe level (PSL) in beampattern synthesis is essential to minimize power leakage from the transmitted beam, thereby preventing interference with non-target users. The constant modulus constants (CMCs) must be considered in SAT beamformer design to enable power amplifiers (PAs) to operate close to the saturation point while remaining in the linear region to achieve maximum efficiency. In this thesis, the uniform rectangular array (URA) beampattern synthesis design problem, formulated to minimize the PSL of the transmit beampattern while satisfying the CMC is studied. A new method called the dynamic points selection (DPS) method is proposed to ensure the suppression of PSL in the continuous spatial domain by finding the local peaks for suppression. The proposed method can reduce the number of constraints, decrease the computation time, and receive lower PSL. For the non-convex constraints, the optimization problem is reformulated to semidefinite programming (SDP) with a rank 1 constraint and is solved by a convex iterative algorithm. Simulation results reveal the advantages of the proposed method for suppressing the lower PSL under CMCs compared to existing methods.