A theoretical analysis of noise in a high-power cascaded fiber amplifier is presented. Unlike the noise theory in low power communication, the noise of a high power system is redefined as the leaked output energy betw...A theoretical analysis of noise in a high-power cascaded fiber amplifier is presented. Unlike the noise theory in low power communication, the noise of a high power system is redefined as the leaked output energy between pulses with coherent beat noise uncounted. This definition is more appropriate for high power usage in which the pulse energy receives more attention than the pulse shape integrity. Then the low power pre-amplifying stages are considered as linear amplification and analyzed by linear theory. In the high-power amplification stages, the inversion is assumed to recover linearly in the time interval between pulses. The time shape of the output pulse is different from that of the input signal because of different gains at the front and back ends of the pulse. Then, a criterion is provided to distinguish the nonlinear and linear amplifications based on the signal-to-noise ratio （SNR） analysis. Then, an experiment that shows that the output SNR actually drops off in nonlinear amplification is performed. The change in the noise factor can be well evaluated by pulse shape distortion.展开更多
基金Project supported by the National High Technology Research and Development Program of China(Grant No.2011AA8042032)the National Natural Science Foundation of China(Grant Nos.61077034 and 61301190)
文摘A theoretical analysis of noise in a high-power cascaded fiber amplifier is presented. Unlike the noise theory in low power communication, the noise of a high power system is redefined as the leaked output energy between pulses with coherent beat noise uncounted. This definition is more appropriate for high power usage in which the pulse energy receives more attention than the pulse shape integrity. Then the low power pre-amplifying stages are considered as linear amplification and analyzed by linear theory. In the high-power amplification stages, the inversion is assumed to recover linearly in the time interval between pulses. The time shape of the output pulse is different from that of the input signal because of different gains at the front and back ends of the pulse. Then, a criterion is provided to distinguish the nonlinear and linear amplifications based on the signal-to-noise ratio （SNR） analysis. Then, an experiment that shows that the output SNR actually drops off in nonlinear amplification is performed. The change in the noise factor can be well evaluated by pulse shape distortion.
