多輸入多輸出之全雙工無線電自我干擾消除

Abstract

2013年，史丹佛大學的一個研究團隊已經成功的建構了單輸入單輸出全雙工無線電系統，2014年他們進一步完成了多輸入多輸出全雙工無線電系統。多輸入多輸出全雙工無線電除了有自干擾，還有交叉干擾的問題要克服。而干擾消除器的數目更會隨著天線數增加而呈指數成長。史丹佛大學團隊發表了多輸入多輸出全雙工無線電自干擾與交叉干擾消除的低複雜度架構，但是並沒有說明自干擾與交叉干擾消除詳細的演算法。本論文旨在開發史丹佛大學低複雜度架構之自干擾與交叉干擾消除演算法。首先我們研究了現有的單輸入單輸出與多輸入多輸出之自干擾與交叉干擾消除演算法，這包含了類比與數位的消除器，接著我們提出了兩種不同的低複雜度演算法來完成自干擾與交叉干擾的數位消除，最後我們建立了一個完整的多輸入多輸出模擬平台來驗證所使用的方法，模擬結果顯示我們所提出的方發可達到系統的要求，即自干擾消除要達102分貝以上而交叉干擾消除要達78分貝以上。

In 2013, a research team in Stanford University announced that it had successfully developed a full-duplex radio (FDR) system. In 2014 it further declared that a MIMO FDR system had also been developed. In MIMO FDR, both self-interference (SI) and cross-interference (CI) cancellation are needed to cancel. Also, the number of the cancellers will increase exponentially as the MIMO dimension is increased. The Stanford team has revealed its low-complexity architectures for SI and CI cancellation, but detailed algorithms for the identification of the digital cancellers remain undisclosed. This thesis aims to study the low-complexity MIMO cancellers. First, identification algorithms for analog and digital cancellers in SISO/MIMO FDR system are studied. Then, two identification algorithms for the low-complexity SI and CI digital cancellers are proposed. Finally, a complete MIMO FDR simulation platform is built to evaluate the performance of proposed methods. Simulations show that the proposed algorithms, along with analog cancellation, can effectively cancel interference and meet the requirement, i.e, 102dB for SI and 78dB for CI.