適應性OFDM系統的設計和分析

Abstract

無線通道因為電磁波的反射、干涉與繞社和都卜勒的效應會產生頻率選擇性與時變性，因而會造成系統的效能低落。適應性調變是一個不錯的方法在解決這個問題上，因此在本篇論文中考慮將適應性調變應用在OFDM系統上的實際問題。從許多論文與期刊上的分析與模擬結果可以知道，適應性調變可以有效地解決通道衰減的問題，但是在真實的環境中，卻存在有通道估測錯誤的問題。因此，論文中我實際以LS演算法探討通道估測錯誤對系統效能的影響。而為了研究雜訊估測誤差對系統的影響，我將Aldana所發明的決策導向雜訊估測做了修改並且將這個演算法實際地用在估測適應性調變OFDM系統的通道雜訊上面。而除了通道資訊的誤差之外，另一個存在適應性調變OFDM系統的問題就是overhead可能佔去實際上可以使用的頻寬。一個最常見的解決方法就是使用group adaptive系統，而通常決定group modulation type的方法都是使用最差的subcarrier的通道資訊。但是這樣的方法卻會降低系統的輸出，因此在這個部分我提出了兩個方法來決定group modulation type以提升系統的效能。而除了group adaptive系統的使用以外，我們也考慮迴授通道資訊的型態，以企圖減低overhead的產生，最後並以系統模擬分析其中的差異。

Due to the reflection, scattering, and diffraction of the EM wave, the wireless channels have the frequency selective and time varying property. Adaptive modulation is a good approach in solving the problem, thus I apply the adaptive modulation onto the OFDM system in this thesis. According to the analysis and the simulation result in many papers, we know that adaptive modulation can effectively mitigate the problem of the fading channel. But in true environment, the channel estimation error exist. Therefore, I use LS algorithm for pilot symbol based channel estimation to analysis the impact of imperfect channel estimation on the adaptive modulation system. And for the purpose of analyzing the impact of the noise estimation error on the AOFDM, I modify the approach submitted by Aldana to estimate the noise in the wireless channel of the AOFDM system. Besides the cannel information error, another problem is the overhead which may reduce the effective available bandwidth. The most popular solution to this question is using group adaptive modulation system, and usually group modulation type is decided by the subcarrier which has the worst channel information. But this method may decreasing the system throughput. Therefore I proposed two method to decide the modulation type in group adaptive modulation system in order to increasing the system performance. In addition to group adaptive modulation system, we also consider the feedback channel information type to reduce overhead. And difference between two channel information type is analyzed.