Memory Co-Nnpd For The Compensation Of Memory Crosstalk And Hpa Nonlinearity
Year: 2011
Type: conference-paper
Abstract: In [1] and [2], authors proposed two efficient crossoverpredistortion schemes which are capable to compensatesimultaneously HPA nonlinearity and crosstalk effects inMIMO systems. The crosstalk model considered in thesepapers was memoryless one. However, memory effects ofcrosstalk can no longer be ignored due to the broadbandtransmitted signal.Then, in this paper, we demonstrate the effect of memorycrosstalk on the Crossover Neural Network Predistorter(CO-NNPD) proposed in [1]. Along, we propose a newcrossover predistortion structure based on this conventionalCO-NNPD which is capable to enhance good performancein MIMO OFDM systems in presence of HPA nonlinearitieswith taken into account the memory effects of crosstalk. TheLevenberg-Marquardt algorithm (LM) is used for neuralnetwork training, which has proven [3] to exhibit a verygood performance with lower computation complexity andfaster convergence than other algorithms used in literature.This paper is supported with simulation results for theAlamouti STBC MIMO OFDM system in terms of Bit ErrorRate (BER) in Rayleigh fading channel.
Source: HAL
DOI: 10.5281/zenodo.42539
Language: English
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2159317631
