Abstract
In this paper, an improved whale algorithm to optimize pump power and wavelengths of second-order fiber Raman amplifier (FRA) has been proposed. Adopted by the strategies of adaptive gradients and reverse learning, the normal whale algorithm is greatly improved. An optimized FRA is designed based on the six-wavelength backward pump structure. Compared by optimizing both wavelength and power with optimizing only power, the gain flatness curves of the two algorithms, the optimal combination scheme, and the corresponding noise characteristics are got. A 6-pumped FRA has been designed in which the error of on-off gain is 0.08% and the ripple is only 0.18dB.
Export citation and abstract BibTeX RIS
Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.