Robustness Improvement of Compact Predistorters in a CO-OFDM System Using Semiconductor Optical Amplifiers - Université de Bretagne Occidentale
Communication Dans Un Congrès Année : 2018

Robustness Improvement of Compact Predistorters in a CO-OFDM System Using Semiconductor Optical Amplifiers

Résumé

This paper focuses on digital signal processing techniques compensating the non-linear effects inherent to optical modulation or amplification within a coherent optical orthogonal frequency division multiplexing (CO-OFDM) transmitter featuring a semiconductor optical amplifier (SOA). Two complementary approaches are specifically considered: crest factor reduction and linearization using a digital baseband predistortion. Two possible implementations are presented for the digital predistortion block. One is the basic solution of static compensation; the other consists in a parallel two-box digital polynomial predistorter (PTB) combining, in parallel, a static nonlinear block with a memory polynomial predistorter. Moreover, we show that a linearization of the dynamic behavior of the transmitter, in addition to peak-to-average power ratio (PAPR) reduction, offers better performance. The robustness of the predistorters or of the combined predistorter/PAPR reduction blocks is assessed for various scenarios. Finally, we propose solutions to improve robustness via multipoint identification or switching strategies for predistortion.
Fichier principal
Vignette du fichier
ICM2017_Mohamad_Younes.pdf (463.66 Ko) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01697205 , version 1 (31-01-2018)

Identifiants

Citer

Mohamad Younes, Mihai Telescu, Noël Tanguy, Chérif El Valid Diouf, Pascal Morel, et al.. Robustness Improvement of Compact Predistorters in a CO-OFDM System Using Semiconductor Optical Amplifiers. 29th International Conference on Microelectronics (ICM), 2017, Dec 2017, Beirut, Lebanon. ⟨10.1109/ICM.2017.8268828⟩. ⟨hal-01697205⟩
340 Consultations
310 Téléchargements

Altmetric

Partager

More