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），將預測範圍提昇至10毫秒以上。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）、可變步階參數最小均方演算法（VSLMS）、與傳統的固定步階參數最小均方演算法（LMS）等調適性濾波技術，並且考慮通道特性以及都卜勒頻率等因素，做調適性傳輸的調適性遠時程瑞雷衰退通道預測演算法設計及其評估，以及比較它們對於高斯雜訊的抵抗能力，並考慮運算複雜度，以評估其實現之可能性。 The goal of this thesis is to discuss how to implement the adaptive long range prediction algorithm (adaptive LRP) practically and extend the prediction range to more than ten milliseconds. Here we are going to design and make the performance evaluation of the adaptive LRP algorithm for the adaptive transmission under the scenario of Rayleigh fading channels by considering three kinds of adaptive filtering techniques：recursive least-square algorithm (RLS), traditional least-mean-square algorithm using a fixed step-size parameter (LMS), and variable step-size least-mean-square algorithm (VSLMS). Furthermore, we will also take account of the effects of channel characteristics and Doppler frequency shift in the design. In the end, the resistance to Gaussian noise and the computational complexity of them will be evaluated.
