正交多頻多工系統使用複雜度資訊之位元寬度決定法

Abstract

我們提出使用複雜度資訊的自動化轉換浮點數到固定點數演算法，演算法的目標是縮小硬體複雜度並且同時減少模擬時間。我們提出的演算法同時考慮到了整數以及小數的位元寬度。對整數的位元寬度來說，演算法決定需要多少位元寬度以避免溢位。對小數位元寬度來說，我們提出的演算法使用上邊界以及下邊界來得到最後的結果。我們把提出來的演算法應用在正交多頻多工系統上面。結果顯示提出來的演算法比循序搜尋法以及複雜度誤差度量測法減少大約30%的摹擬時間。

An automatic algorithm using complexity information for the floating point to fixed point conversion is proposed. The goal of the proposed algorithm is to minimize the hardware complexity and reduce the simulation times. The algorithm considers both the integer bit width and the fraction bit width. For the integer bit width, the algorithm identifies numbers of the integer bit width to prevent the overflow. For the fraction bit width, the algorithm uses the lower bound and the upper bound to find the results. We apply the proposed algorithm to the OFDM system. The results show that the proposed algorithm reduces almost 30% simulation time than complexity-and-distortion measure and sequential search method.