Comparison of nonlinear coefficient of optical fibers at 1550 nm

Abstract

ThA3FOM W-shaped DCF. Fig. 1. Attenuation and dispersion spectra of high s 8 "E 0 N E Dispersion (pdnmlkm) ThA3 Fig. 2. Dispersion value versus n2 value.were demanded for DCF. Large FOM was requested to decrease optical signal-to-noise ratio (SNR). High negative dispersion was required to shorten the necessary length for compact packaging. Table 1 shows characters of DCFs optimized for above requirements. High negative dispersion (-163 ps/nm/km) and high FOM (323 ps/nm/dB) were realized with W-shaped index profiles.Recently, WDM technique has been studied actively for higher bit-rate transmission such as 10 Gbit/s system etc. For this technique, slope compensation and suppressed nonlinearity were desired. In order to discuss the flat compensation at a wide wavelength range, a dispersion slope compensation rate, shown with the Eq. (l), is proposed,where slope was defined as the difference of dispersion values from 1.53 pm to 1.56 pm divided by a wavelength interval. The nearer to 100% this rate is, the higher the compensating efficiency is. Table 1 also shows the compensating rate of DFCF. Because matched clad DCF could not give negative slope, Wshaped DCF was suitable for WDM transmission. On the WDM transmission, optical power density becomes very large because plural optical signals are inputted into a fiber at the same time. Then a pulse is distorted by nonlinear phenomenon (e.g., cross-phase modulation, XPM; four-wave mixing, FWM; and stimulated Brillouin scattering, SBS), therefore suppression of nonlinearity is important. XPM is the thirdorder nonlinear phenomenon and can be estimated from nonlinear refraction rate (n,) of fiber? We measured nz values of W-shaped DCF and matched clad DCF using XPM method. Figure 2 shows the relation of dispersion values versus n, values of each profile. The n2 values of W-shaped DCF were smaller than those of matched clad DCFs at the same dispersion value, which was caused by the higher negative dispersion of W-shaped profile. In this point of view, W-shaped DCF is also better than matched clad DCF. FWM was not observed due to the large dispersion values of both SMF and DCF. Because threshold value of SBS was bigger than XPM, suppression of XPM was equated with that of SBS.It is concluded that W-shaped index profile is suitable for DCF and adequate for WDM transmission from the following points of view: (1) low nonlinearity, (2) compact module, (3) flat compensation. 1. Y. Akasaka et al., in OFC'95 Technical Digest, paper ThH3. 2. Y. Akasaka et al., in Proc. ECOC'95, paper We.B.2.4. 3. R. H. Stolen et al., in OFC'95 Technical Digest, paper FD1.In this paper we report on measurements of the nonlinear coefficient n2/Aeff (nonlinear refractive index/effective area) of different fibers of interest in transmission systems at 1550 nm: large-effective area fiber dispersion-shifted fiber (LEAF),' dispersion-compensation fiber (DCF), and standard dispersionSynthesizer

Effects of self-phase modulation on sub-500 fs pulse transmission over dispersion compensated fiber links

“…The input pulse FWHM is 400 fs (solid line). reported values of other DCF's [26], [27]. These simulation parameters were confirmed by matching the simulated output spectra from either the SSMF (2.06 km) or the DCF (0.44 km) alone with the corresponding measured spectra, respectively.…”

“…We noted that our simplified simulation includes only the second-order dispersion and the SPM term, where the parameters of DCF are the same as those of SSMF except the sign of the dispersion. Usually, the ratio of nonlinear coefficient to fiber dispersion is less for real DCF than for SSMF [26], [27]. Therefore, our simulation is conservative and overestimates the effect of SPM in the DCF.…”

Comparative study of interferometric and noninterferometric measurement schemes for fiber nonlinear coefficient measurement

“…So, the measurements are made in fibers that are short enough to neglect loss and dispersion. In SPM method with pulsed laser the spectral broadening factor for a Gaussian pulse is given by [13], [14] ( ) ( )…”