Abstract

Of all the lasers in use today, it is only the semiconductor laser where the quantum noise, inevitable in all lasers, is of major practical concern. This is due to two reasons. The first one is that since semiconductor lasers have a very small physical volume, the cavity quality factor is very small, which results, according the the Schalow-Townes formula, in a large spectral width of the laser output. The second reason is that due to the widespread application of semiconductor lasers in real-life optical fiber comunications systems, the noise aspects are of real concern. The basic physical mechanism of this noise is the addition of spontaneous emission power to the laser field. This effect is compounded and amplified by coupling between intensity and phase fluctuations, which are characteristic of semiconductor lasers. In the talk we will review those physical effects and show some possible ways of reducing the laser linewidth. These ways include the use of large external cavity and/or coupling of frequency dependent loss mechanism into the laser oscillators. The effect of spontaneous emission is also a limiting factor in optical fiber amplifier systems, which are assuming major practical importance. In this case the major noise mechanism is beating between the signal propagating in the fiber and the background amplified spontaneous emission power. The basic limitation on signal to noise ratio in a practical system because of this mechanism will be reviewed. Experiments performed in various laboratories will be discussed and compared to the theory. A limiting noise behavior of the optical amplifier fiber system is shown to consist of a transmission system with uniformly distributed gain along the whole length of the fiber. The penalty factors for using other schemes of optical fiber amplification compared to the ideal system will be discussed. The last mechanism to be discussed will be that of double Rayleigh scattering in fibers. This converts phase noise to intensity noise. The resulting limiting RIN will be discussed and compared with new experiments.

© 1990 Optical Society of America