本論文旨在探討如何真正地實現調適性遠時程通道預測演算法(Adaptive Long Range Prediction,簡稱Adaptive LRP),並且將此結果應用在調適性調變系統上。本論文將針對在多重路徑衰退通道之中,使用遞迴最小平方演算法(RLS)、傳統的固定步階參數最小均方演算法(LMS)與可變步階參數最小均方演算法(VSLMS)等調適性濾波技術,並且考慮通道特性等因素,做調適性調變系統的調適性遠時程瑞雷衰退通道預測演算法設計及其評估,以及比較它們位元錯誤率和BPS性能,並考慮運算複雜度,以評估其實現之可能性。再者,根據不同的都卜勒頻率偏移決定適合的遠時程通道預測演算法階數及調適性調變系統的臨界值。 The aim of this thesis is to propose three kinds of adaptive long rang prediction (Adaptive LRP) algorithms and extend the prediction result to the adaptive modulation. The adaptive LRP algorithms we used here are recursive least-square (RLS) algorithm, traditional least-mean-square (LMS) algorithm using a fixed step-size parameter and variable step-size least-mean-square (VSLMS) algorithm. We will investigate the BER and BPS performance of adaptive modulation in Rayleigh fading channel by using our proposed adaptive LRP algorithms and perfect prediction. Furthermore, we will also consider different speed zone to decide the suitable order of the adaptive LRP algorithms and the threshold of the adaptive modulation system.