較佳錯誤效能的多輸入多輸出線性接收機的訊號系統

Abstract

線性接收機是一個簡單的方法用於無線多輸入多輸出(MIMO)通訊系統的解調而且又屬於低複雜度。最大概率(ML)接收機一方面來說擁有最佳的效能，但是複雜度卻非常的高。在這篇論文中，我們提出在多輸入多輸出雷利(Rayleigh)衰減通道情形中的多樣性和多工制衡(DMT)的分析而且使用一般的高斯輸入及線性接收機。論文顯示出線性接收機的DMT被低估了。我們藉由改變高斯輸入端的協方差矩陣(covariance matrix)並且使用時間編碼橫跨多次通道使用次數，DMT效能可以很大的被改善而且在固定的傳輸速度之下線性接收機可以達到和ML接收機一樣的多樣性(diversity)；因此此方法遠遠優於目前廣泛採用獨立同高斯分佈輸入。此外，停運(outage)機率和錯誤機率在高訊雜比(SNR)的情況之下都會被改善。我們也提供了數值模擬來驗證我們的結果。

Linear receiver is an easy way for demodulations in wireless multiple-input multiple-output (MIMO) communications due to its low complexity. The ML receiver on the other hand, though having optimal performance, has very high complexity. In this thesis, DMT analysis of MIMO Rayleigh fading channels using general colored Gaussian inputs and linear receivers are presented. It is shown that the DMT of linear receivers is in general underestimated. We propose that by varying the rank of the covariance matrix of the Gaussian input and allowing temporal coding across multiple channel uses, the DMT performance can be significantly improved such that the linear receivers can achieve the same diversity as the ML receivers at a fixed data rate; hence it is much better than the widely adopted independent and identically distributed Gaussian input. Furthermore, the outage probability and the error probability can be improved in the high signal-to-noise-ratio (SNR) regime. Numerical simulations are also provided to justify our claims.