It has been shown that saturated EDFAs undergo large and rapid per-channel output power transients when wavelength channels are added and dropped. This effect has stimulated the development of gain-clamped or gain-stabilized EDFAs. Several methods for gain clamping have been proposed and demonstrated. Among these are methods which gain clamp each EDFA and methods which gain clamp a chain of EDFAs by inserting a compensating channel which is propagated to and gain clamps all subsequent amplifiers. Both of these methods have been shown to work well in networks where only one gain clamping technique is used. To foster network interoperability, it is necessary either that the different types of amplifiers be compatible or that some means is found to make them compatible. Here we use wavelength-domain simulation to show that these two gain stabilization techniques are intrinsically incompatible. We also show how simple modifications can make them compatible.
[1]
R.W. Tkach,et al.
Fast-link control protection of surviving channels in multiwavelength optical networks
,
1997,
IEEE Photonics Technology Letters.
[2]
V. Havard,et al.
Upper limit and control scheme for power per channel in optically-amplified WDM systems
,
1996,
Proceedings of European Conference on Optical Communication.
[3]
A. Srivastava,et al.
EDFA transient response to channel loss in WDM transmission system
,
1997,
IEEE Photonics Technology Letters.
[4]
A. Von Lehmen.
Large-scale demonstration of multi-wavelength optical networking in MONET Washington DC network
,
1999
.