用於時變通道下有限回饋傳送波束形成之基於吞吐量模式選擇準則

Abstract

在本篇論文中，吾人考慮適用於時變通道環境之多輸入單輸出 （Multiple-Input Single-Output, MISO）有限回饋預編碼 (Limited Feedback Precoding) 系統的雙模式編碼方法，吾人持用之通道模式係由一階馬可夫通道模型描述。因為通道統計特性隨著時間碼框改變，系統需提供不同大小的碼書以提升性能表現。在吾人的系統中，採用的兩種不同的碼書係遵循IEEE 802.16e 標準之定義，並分別對應到不同的通道向量量化程度。基於公平原則，系統中的雙模式具有相同的平均回傳率。為了在不同的通道相關性下選擇適當的碼書，吾人提出一個以吞吐量為準則的模式選擇方法協助接收機決定使用的碼書類型。吾人亦提出實際可行的準則實現模式及選擇方法供接收機使用。模擬顯示吾人所提出的模式選擇準則在套用於雙模式系統中可以達到比單模式系統更好的系統性能表現，並可與現有的模式選擇準則相比較。

In this thesis, we consider a dual-mode precoding method for limited-feedback multiple-input single-output (MISO) systems over time-varying channels, which follows the first-order Markov channel model. Since the channel statistics varies over time frames, codebooks of different sizes should be available to improve the system performance. In the considered system, there are two codebooks corresponding to different quantizations of channel vectors as adopted in the IEEE 802.16e standard. For fairness, the two modes have the same feedback rates. To choose one suitable mode for use under different channel correlations, we propose a throughput-based mode selection criterion to help the receiver decide the codebook to be used. A practical way to implement the criterion at the receiver is also proposed. Simulations indicate that the proposed scheme is superior to that with a single codebook and perform comparably to those existing mode selection criteria.