全雙工無線電之自我干擾消除

Abstract

全雙工無線電可增加一倍的通訊系統容量，而且它最近也吸引了相當多研究團隊的注意，在2013年，史丹佛大學的一個研究團隊宣稱他們已經成功的建構了全雙工無線電系統，如眾所皆知的，全雙工無線電最關鍵的問題就是自干擾消除，史丹佛大學團隊已經發表了自干擾消除的架構，但是並沒有說明自干擾消除詳細的演算法，因此，本篇論文旨在史丹佛大學的自干擾消除架構下開發自干擾消除演算法，據發現，史丹佛大學的架構需要三部分的消除演算法，即是類比消除器、量化類比消除器，以及數位消除器。我們提出了幾個不同的演算法來完成自干擾消除的動作，模擬結果顯示自干擾消除程度可達到要求，即110分貝以上。

Full duplex radio (FDR) can double the capacity of a communication system, and it attracts much research attention recently. In 2013, a research team in Stanford University announces that it has successfully developed a FDR system. As well known, the key problem in a FDR system is the self-interference (SI) cancellation. The Stanford team has revealed its architecture of SI cancellation. However, detailed algorithms for the identification of the cancellers remain undisclosed. This thesis aims to develop the identification algorithms for the SI cancellers in Stanford’s SI architecture. It is found that three kinds of identification algorithms are needed in Stanford’s SI architecture, i.e, for analog canceller, quantized analogy canceller, and digital canceller. We have proposed various algorithms to do the identification work. Simulations show that the cancellation level yielded by the proposed algorithms can meet the requirement, i.e, larger than 110dB.