论文部分内容阅读
由于格型结构预测器所具有的优点,使它在许多自适应系统中得到了很好的应用。本文着重研究了采用归一化 SG 算法的自适应格型预测器。文中给出了归一化算法的推导,讨论了预测器系数的收敛特性以及格型预测器级数对预测器的影响。
Due to the advantages of the lattice structure predictor, it has been well applied in many adaptive systems. This article focuses on adaptive lattice predictors using normalized SG algorithms. The derivation of the normalization algorithm is given in this paper. The convergence of the predictor coefficients and the influence of the predictor progression on the predictor are discussed.