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