高自由度之非均勻線性天線陣列架構搜尋演算法

Abstract

近年來，由於非均線性陣列(Non-uniform Linear Array) 架構具備提升智慧型天線陣列內隱含自由度的潛能，十分引人注目。 額外的自由度給予非均勻線性陣列偵測超出其天線數量的入射信號的能力；而當使用均勻線性天線陣列時，能夠被偵測到的信號數量必定小於陣列的天線數。因為在新標準中對大數量天線的使用有興趣，使得非均勻線性陣列對即將來臨的第五代無線通訊標準來說，是極有吸引力的。本論文提出一個基於間距限制陣列找尋非均勻線性陣列的演算法，這種改善的非均勻線性陣列能夠提供較目前現存架構都還要高的隱含自由度。 利用接收信號的二階統計特性，我們從提出的陣列所得到的自由度得以發揮用途。 不只如此，空間平滑(Spatial Smoothing)和陣列內差(Array Interpolation)的混合方法應用在提出的天線陣列上顯示極佳的效能表現，即使提出的陣列不具備范德蒙 (Vandermonde)結構。模擬結果顯示我們的架構在訊號到達方向估測的方均根誤差和接收波束成型的訊號對干擾雜訊比等應用上表現優於大部分的現存非均勻線性架構。

The design of non-uniform linear array (NLA) configurations have drawn plenty of attention in recent years due to their potential for increasing obtainable degrees of freedom (DOF) in smart antenna arrays. The extra DOF give NLA the ability to detect more number of signals than that of the sensors in the array while detectable number of sources must be smaller than that of antennas when uniform linear array (ULA) configuration is applied. That makes NLA an appealing structure for the upcoming 5G wireless communication standard for the reason that the applications of a massive amount of antennas are of great interest in the new standard. This thesis proposes an algorithm of finding an improved non-uniform linear array configuration based on spacing-constrained array, which is able to provide higher obtainable DOF than existing NLA configurations. We exploit the second-order statistics of the received signals to make use of the additional DOF we gain from the proposed structure. Furthermore, a mixture of spatial smoothing and array interpolation is applied to show that great performance can be achieved even without having Vandermonde structure. Simulation results show that our structure outperforms most of the existing structures in applications such as DOA estimations and receive beamforming in terms of root mean square error (RMSE) and signal to interference-plus-noise ratio (SINR), respectively.