30 IEEE PHOTONICS TECHNOLOGY LETTERS, VOL. 14, NO. 1, JANUARY 2002
Study of the Operating Regime for All-Optical
Passive 2R Regeneration of Dispersion-Managed RZ
Data at 40 Gb/s Using In-Line NOLMs
Sonia Boscolo, Sergei K. Turitsyn, Associate Member, IEEE, and Keith J. Blow
Abstract—In this letter, we numerically demonstrate that the
use of inline nonlinear optical loop mirrors in strongly dispersion-
managed transmission systems dominated by pulse distortion and
amplitude noise can achieve all-optical passive 2R regeneration of a
40-Gb/s return-to-zero data stream. We define the tolerance limits
of this result to the parameters of the input pulses.
Index Terms—Dispersion management, nonlinear optical loop
mirrors, passive regeneration.
I. INTRODUCTION
T
RANSMISSION of return-to-zero (RZ) dispersion-man-
aged (DM) pulses is a key technology in high bitrate op-
tical transmission systems that manages fiber nonlinearity and
suppresses interchannel crosstalk in wavelength-division multi-
plexed systems [1], [2]. However, it has been observed that in-
trachannel nonlinear effects can be significant for strongly DM
systems, leading to serious penalties such as amplitude jitter of
the main signals and the “ghost” pulse generation at the zero bits
via energy transfer between signals [3], [4].
In this letter, we study the use of inline nonlinear optical loop
mirrors (NOLMs) as a general technique for all-optical passive
2R regeneration of RZ data in high-speed strongly DM sys-
tems, where the transmission performance is mainly degraded
by pulse distortion and amplitude noise. Indeed, as a NOLM
introduces no temporal reference point into the system, it does
not improve the timing jitter in the system. On the other hand,
loop mirror intensity filtering allows for partial regeneration of
pulse amplitude and shape [5]. Recently, we have demonstrated
a feasibility of 40-Gb/s stable soliton transmission over unlim-
ited distances in standard fiber [6]. Here, we investigate the tol-
erance of this result.
II. SYSTEM DESCRIPTION
As a sample system for demonstration of the technique, we
used a symmetric dispersion map [7] (see Fig. 1). The transmis-
sion line is composed of an equal number of 32.3-km standard
monomode fiber (SMF) and 6.8-km dispersion compensating
fiber (DCF). The map consists of an alternation of two different
fiber combinations: a SMF-DCF and a DCF-SMF block. The
Manuscript received April 30, 2001. S. Boscolo was supported by the U.K.
Engineering and Physical Sciences Research Council (EPRSC).
The authors are with the Photonics Research Group, School of Engineering
and Applied Science, Aston University, Birmingham B4 7ET, U.K. (e-mail:
boscolsa@aston.ac.uk).
Publisher Item Identifier S 1041-1135(02)00069-1.
Fig. 1. Schematic diagram of one element of the periodic transmission system.
dispersion coefficients are 15.0 ps/(nm km) for the SMF and
71.2 ps/(nm km) for the DCF. The effective area is 70 m
for the SMF and 30 m for the DCF. The attenuation is 0.22
dB/km in the SMF and 0.65 dB/km in the DCF. A combination
of an erbium-doped fiber amplifier (EDFA) and a fixed Gaussian
filter follows each of the two blocks. The amplifier has a noise
figure of 4.5 dB and a power gain of 11.9 dB, and the bandwidth
of the filter is taken to be 1.5 times the bandwith of the pulses
used. The NOLM is placed into the transmission line every five
periods of the dispersion map (as shown in Fig. 1). The NOLM
incorporates a 50 : 50 coupler, and a loop of dispersion-shifted
fiber (DSF) with zero dispersion, an attenuation of 0.3 dB/km,
and an effective area of 25 m . Unbalancing of the NOLM is
achieved with an asymmetrically placed EDFA close to the loop
coupler. The continuous-wave (CW) input–output power map-
ping of the NOLM is given by
(1)
where
and output and input powers, respectively;
power gain of the loop amplifier;
nonlinear coefficient of the loop fiber;
where is the loop length, and is the loss co-
efficient [km ] of the loop fiber; and .
We use here a simple procedure to estimate the NOLM’s charac-
teristics needed to adapt the NOLM to the DM line. First, we fix
a suitable value for the loop length, km throughout this
letter. Then, from the condition of stable operational regime, we
calculate the loop gain for a given input power . Using this
value of in (1), we calculate , which is greater
than 1 in the range of used in our simulations, and we re-
store the peak power of pulses after the NOLM by the use of an
attenuator with a power loss of .
III. TRANSMISSION SIMULATIONS AND RESULTS
First, single chirp-free Gaussian-shaped pulses are launched
into the system with different pulsewidths and peak powers.
1041–1135/02$17.00 © 2002 IEEE