DocumentCode :
895798
Title :
Novel shooting algorithm for highly efficient analysis of fiber Raman amplifiers
Author :
Han, Qun ; Ning, Jiping ; Zhang, Huayong ; Chen, Zhiqiang
Author_Institution :
Coll. of Precision Instrum. & Optoelectron. Eng., Tianjin Univ., China
Volume :
24
Issue :
4
fYear :
2006
fDate :
4/1/2006 12:00:00 AM
Firstpage :
1946
Lastpage :
1952
Abstract :
An efficient and stable algorithm based on the simple-shooting method to solve the fiber Raman-amplifier (FRA) equations is proposed and numerically demonstrated. In the algorithm, the conventional Newton-Raphson method is substituted by a more efficient and stable modified Newton method to serve as the correction mechanism in the shooting iterations. By introducing the Broyden´s rank-one method into the modified Newton method, the time-consuming calculation of the Jacobian matrix is dramatically relieved, and the efficiency of the proposed algorithm is further remarkably improved. An effective initial-guess-providing technique for the shooting method has also been proposed. Simulation results show that the efficiency of the proposed algorithm is improved more than three times compared with those of the previously reported methods. This algorithm can be used in the design of various kinds of FRAs to predict their gain and noise performance accurately.
Keywords :
Newton method; Raman lasers; initial value problems; optical fibre amplifiers; Broyden rank-one method; Jacobian matrix; fiber Raman amplifiers; initial-guess-providing technique; modified Newton method; shooting iterations; Algorithm design and analysis; Equations; Jacobian matrices; Newton method; Optical fiber amplifiers; Optical fiber communication; Raman scattering; Stability; Stimulated emission; US Department of Transportation; Boundary-value problems; Raman scattering; fiber Raman amplifier (FRA); modeling;
fLanguage :
English
Journal_Title :
Lightwave Technology, Journal of
Publisher :
ieee
ISSN :
0733-8724
Type :
jour
DOI :
10.1109/JLT.2006.871008
Filename :
1618786
Link To Document :
بازگشت