並列処理を用いたブラインド等化

概要

論文の詳細を見る
In a transmission system, the deconvolution is the method of estimating an inverse system of a channel (equalizer) only from received signals and recoverring send signals. These representative of on-line blind deconvolution algorthms are the probability distribution congruence method and the minimum entropy method. In these on-line algorthms, when a cost function for estimating an inverse system of a channel is optimized, coefficients of an inverse system are estimated. In these methods, the design of these cost functions is used from the information of the probability distribution for send signals. But generally, the probability distribution of send signals is unknown. Therefore, we assume a probability distribution from send signals. If the assumed probability distribution of send signals does not correspond to the truth probability distribution, recoverd signals do not give a precise estimation. First, we estimate the probability distribution of send signals only from received signals. Next, we design the cost function using information of the probability distribution for estimated send signals. We confirm the estimation through computer simulation.
前橋工科大学の論文
1998-04-09