Abstract
The trans-Atlantic cable to be laid in 1995 will not use solitons. Instead, it will use a dispersion shifted fiber, with zero dispersion at and near the 1.5 μ wavelength of the Erbium amplifier. The transmission will be “linear,” at 1.5 μ, suppressing effects of the nonlinearity of the fiber by staggering slightly positive and slightly negative dispersive fiber segments. This design was chosen, rather than the repeaterless soliton transmission proposed by Hasegawa[1], and experimentally demonstrated by Mollenauer[2], because of an effect characteristic of amplified solitons, the so called Gordon-Haus effect[3]. This effect is due to the shift of the soliton carrier frequency experienced when a soliton is amplified by an amplifier with spontaneous emission noise. The carrier frequency performs a random walk. Even though the “steps” are small, the cumulative effect of travel over long distances shifts the timing of the solitons, leading to errors. Transmission over trans-Atlantic distances is not prevented by this effect. However, the effect increases with increasing power and sets an upper limit to the signal power. A lower limit on the signal power is set by the standard ratio of signal power to noise power, the S/N ratio. The window of permissible power, for a given bit-error rate, is too narrow to assure long term reliability.
© 1993 Optical Society of America
PDF ArticleMore Like This
C. R. Mirasso, L. Pesquera, A, ., and Mecozzi
TuB.4 Nonlinear Guided-Wave Phenomena (NP) 1993
S. Wabnitz
WM4 Optical Fiber Communication Conference (OFC) 1996
N. J. Doran and W. Fovysiak
CWF7 Conference on Lasers and Electro-Optics (CLEO:S&I) 1993