Quality surveillance algorithm for erbium-doped fiber amplifiers

Gain-flattened erbium-doped fiber amplifiers (EDFAs) have become key components for optical communication systems and networks. A failure of one of these devices interrupts data transmission on the affected link completely and causes significant loss of income to the carrier. Therefore, detection of the EDFA's performance degradation is mandatory, since amplifiers can be replaced before they fail. State of the art techniques can only detect aging of the pump laser. In this paper, a novel algorithm is presented that allows for the detection of various degradation mechanisms. The algorithm allows to supervise the complete optical path within the EDFA and is based on appropriate modeling of the gain characteristics of the erbium-doped fiber. Thus, aging effects can be detected for arbitrary channel loads and power distributions at the input of the amplifier stage. Since the effect of pump excited state absorption (ESA) is included in the model, the algorithm is also applicable to amplifier stages using high power pumps.