基於軟式解調輸出球面解碼法之效能改進

Abstract

在多重輸入多重輸出(MIMO)的無線通訊系統中，已有許多偵測信號的方法被提出，像是強制歸零(zero-forcing)檢測法、最小均方誤差(minimum mean square error)檢測法、最大概度(maximum likelihood)演算法、以及球面解碼(sphere decoding)演算法等。這些方法要不就是無法產生軟式輸出(soft output)，不然就是產生軟式輸出需要大量的運算，因此實用上皆有其限制。 本篇論文中，我們主要著重在軟式解調輸出球面解碼演算法與其降低運算複雜度的方法。簡言之，我們提出一種新的軟式輸出演算法，相較於用於降低球面解碼法複雜度的傳統方法，此新方法可以用較少的運算量來達到較佳效能。

Many solutions for detecting signals transmitted over flat-faded multiple input multiple output (MIMO) channels have been proposed, e.g., the zero-forcing (ZF) detector, minimum mean squared error (MMSE) detector, brutal-force maximum likelihood (ML) detector, sphere decoding (SD) algorithm, to name a few. These approaches however either do not provide soft-output or suffer from high complexity when being modified to a soft-output counterpart.

In this thesis, we focus on the soft-output SD algorithm and the methods of its computational complexity reduction. We will first illustrate that the single tree search (STS), ordered QR decomposition, channel matrix regularization, and log-likelihood ratio clipping can reduce the decoding complexity but at a price of the performance degradation. We then present a new algorithm for soft detection in an MIMO system. Simulations show that our new method can achieve simultaneously less performance degradation and less decoding complexity than the existing methods.