空間調製多輸出輸入系統之低複雜度接收機

Abstract

近年來，為了提升多天線之系統效能，研究者發展出空間調製(Spatial Modulation, SM)的技術，如空間位移碼(Space Shift Keying, SSK)、廣義空間調製(Generalized Spatial Modulation, GSM)及空間調製多輸入多輸出系統(Spatially Modulated-MIMO, SM-MIMO)等。在本篇論文中，我們考慮SM-MIMO系統之低複雜度接收機設計。眾所週知，最佳解最大相似(Maximum Likelihood, ML)接收機之運算複雜度極高，因此我們提出了兩種演算法來解決複雜度的問題，即高斯近似法及QR投影法，主要的想法是把訊號偵測分為兩階段，天線索引及傳送訊號符元偵測，並且運用一些維度降低的方法。這兩種方法可擴展到具通道編碼之SM-MIMO系統軟性位元計算。模擬結果顯示，我們所提出的演算法之效能非常接近ML接收機，且可以降低90%的運算複雜度，有利於對SM-MIMO系統中的實際應用。

Recently, spatial modulation (SM) has been developed to improve the performance of multi-antenna systems. Several SM techniques, for example, space shift keying, generalized spatial modulation, and spatially modulated-MIMO (SM-MIMO), have been proposed. In this thesis, we consider low-complexity receiver designs in SM-MIMO systems. As well-known, the optimum receiver is the maximum likelihood (ML) receiver. However, also well-known, the computational complexity of the ML receiver can be prohibitively high. In this thesis, we propose two new algorithms, referred to as Gaussian approximation and QR-projection, to solve this problem. The main idea is to split the detection into two-stage, antenna index and symbol, and apply some dimension reduction techniques. There two methods are also extended to conduct soft-demapping in coded SM-MIMO systems. Simulations show that the performance of the proposed methods is similar to that of the ML receiver, while the computational complexity can be reduced by 90%, facilitating the real-world applications of the SM-MIMO systems.