Premium
Time‐domain simulation of power transients in Raman fibre amplifiers
Author(s) -
Karásek M.,
Kaňka J.,
Honzátko P.,
Peterka P.
Publication year - 2004
Publication title -
international journal of numerical modelling: electronic networks, devices and fields
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.249
H-Index - 30
eISSN - 1099-1204
pISSN - 0894-3370
DOI - 10.1002/jnm.531
Subject(s) - discretization , amplifier , time domain , signal (programming language) , power (physics) , channel (broadcasting) , upstream (networking) , amplified spontaneous emission , physics , control theory (sociology) , computer science , optics , telecommunications , mathematics , mathematical analysis , laser , control (management) , bandwidth (computing) , quantum mechanics , artificial intelligence , computer vision , programming language
We study power transients in Raman fibre amplifiers (RFA). Transients are defined as the output signal power response to abrupt change in the number of signal channels at the input to the RFA (channel addition/removal). The simulations are based on application of a large signal numerical model which incorporates time variation effects, downstream propagation of multiple signals, upstream propagation of pump and both downstream and upstream propagation of amplified spontaneous emission (ASE). System of coupled non‐linear differential equations describing the propagation of the signal, pump and ASE waves along the RFA and their evolution in time represents a two‐boundary value problem. Due to the backward propagating ASE and counter‐directional pumping, an iterative forward and backward solution of propagation equations must be used in order to achieve a steady‐state distribution of signals, pumps, and ASE powers along the RFA. We have used either the fourth‐order Runge–Kutta routine, or an alternative average power analysis (APA) approach to obtain the steady‐state optical power distribution along the fibre. Direct integration is used to obtain time evolution of optical powers as a response to channel addition/removal. Stability of the numerical solution depends on the relation of discretization steps in space and time. Gain locking of the RFA via electronic feedforward pump control derived from a monitoring channel output power has been introduced in the model to study the possibility of suppressing the surviving channel power fluctuations. Copyright © 2004 John Wiley & Sons, Ltd.